HP P4000 SAN Solution User Guide Abstract This guide provides information for configuring and using the HP SAN Solution. It includes hardware configuration and information about designing and implementing a P4000 SAN. The intended audience is system administrators responsible for implementing, maintaining, and managing a P4000 SAN Solution.
© Copyright 2009, 2012 Hewlett-Packard Development Company, L.P. Confidential computer software. Valid license from HP required for possession, use or copying. Consistent with FAR 12.211 and 12.212, Commercial Computer Software, Computer Software Documentation, and Technical Data for Commercial Items are licensed to the U.S. Government under vendor's standard commercial license. The information contained herein is subject to change without notice.
Contents 1 Getting started.........................................................................................12 Using the CMC......................................................................................................................12 Layout of the CMC.............................................................................................................12 Auto discover systems in the CMC........................................................................................
RAID devices by RAID type.................................................................................................34 Planning the RAID configuration...............................................................................................34 Data protection..................................................................................................................34 Using disk RAID with Network RAID in a cluster.....................................................................
Editing a DNS server..........................................................................................................73 Editing a domain name in the DNS suffixes list......................................................................73 Removing a DNS server......................................................................................................73 Removing a domain suffix from the DNS suffixes list................................................................73 Setting up routing..
Enabling SNMP agents.......................................................................................................93 Adding SNMP traps...........................................................................................................95 Using the SNMP MIBs........................................................................................................96 Running diagnostic reports.......................................................................................................
9 Using specialized managers....................................................................121 Failover Manager.................................................................................................................121 Planning the virtual network configuration...........................................................................121 Failover Manager requirements..........................................................................................
Ongoing capacity management.............................................................................................148 Number of volumes and snapshots.....................................................................................148 Reviewing SAN capacity and usage...................................................................................148 Measuring disk capacity and volume size...........................................................................
Shared versus individual characteristics...................................................................................183 Clone point..........................................................................................................................185 Shared snapshot ..................................................................................................................187 Creating SmartClone volumes.......................................................................................
Understanding the performance statistics.............................................................................214 Monitoring and comparing multiple clusters.............................................................................216 Performance monitoring and analysis concepts.........................................................................216 Workloads......................................................................................................................216 Access type.....
Establishing a terminal emulation session on a Windows system.............................................235 Establishing a terminal emulation session on a Linux/UNIX system..........................................235 Opening the Configuration Interface from the terminal emulation session.................................236 Logging in to the Configuration Interface.................................................................................236 Configuring administrative users..........................
1 Getting started Welcome to the SAN/iQ software and the Centralized Management Console (CMC). Use the CMC to configure and manage the HP P4000 SAN Solution. This product guide provides instructions for configuring individual storage systems, as well as for creating storage clusters, volumes, snapshots, and remote copies.
Navigation window The left vertical pane displays the architecture of your P4000 SAN. Access the Configuration Summary and Best Practice Summary in the navigation window, as well management groups, clusters and storage volumes. • Many items have right-click menus. • Each item in the navigation window has an icon depicting what type of item it is. A faded-looking icon indicates a remote item. Find a description of all the icons used in the CMC in Help→Graphical Legend.
• Tasks—Lets you access all storage configuration tasks. The tasks in this menu are grouped by logical or physical items. Tasks are also accessible through right-click menus and from the Tasks button in the tab window. • Help—Lets you open the online help, access the online upgrades feature, and set CMC preferences. The Graphical Legend is a key to the icons used in the CMC and is available from the Help menu.
Using views and layouts The views and layouts differ for each element of your network that uses the map view.
Figure 2 Default naming conventions for snapshots and SmartClone volumes Changing naming conventions Change the elements that use a default naming convention or change the naming convention itself. If you use the given defaults, the resulting names look like those in Table 2 (page 16). Notice that the volume name carries into all the snapshot elements, including SmartClone volumes, which are created from a snapshot.
• Install the HP P4000 CMC software on a management workstation or server that can connect to the storage systems on the network. • Install an iSCSI initiator, such as the latest version of the Microsoft iSCSI Initiator, on the application server(s). Finding storage systems When you open the CMC, it searches for systems using Auto Discover by Broadcast. If no systems are found, the Find Systems window opens. Add individual IP addresses to find systems.
Figure 3 The SAN/iQ software storage hierarchy 1. Management group 2. Cluster 3. Volume To complete this wizard, you will need the following information: • A name for the management group.
Finding storage systems after the first time The Find settings from your first search are saved in the CMC. Every time you open the CMC, the same search automatically takes place, and the navigation window is populated with all the storage systems that are found. Control the systems that appear in the CMC based on criteria you set in the Find Systems window, opened from either the menu bar or the Getting Started Launch Pad.
• Extremely high network traffic to and from the storage system. • The IP address could have changed if the storage system is configured to use DHCP (not recommended). • The storage system may have been rebooted and is not yet online. • Power could have failed to a network switch that the storage system is connected to. • The CMC might be running on a system that is on a different physical network than the storage system.
2 Working with storage systems Storage systems displayed in the navigation window have a tree structure of configuration categories under them. The storage system configuration categories include: • Diagnostics • Storage • TCP/IP Network Storage system configuration categories Storage system configuration categories allow access to all the configuration tasks for individual storage systems. You must log in to each storage system individually to configure the functions of that storage system.
Automatic login Once you have logged in to a management group, additional log ins are automatic if the same user names and passwords are assigned. If management groups have different user names or passwords, then the automatic log in fails. In that case you must log in manually. 1. Enter the correct user name and password. 2. Click Log In. Logging out of a storage system 1. 2. Select a storage system in the navigation window. Right-click, and select Log Out.
Powering on or off, or rebooting storage systems with modular components Some storage systems are comprised of modular components, that may include: • Disk enclosure • Server blades enclosure • System controller The P4800 is an example of a storage system that is comprised of modular components. Power on the components in the following order: 1. Disk enclosure. 2. Server blades enclosure or system controller.
4. 5. Select Reboot to perform a software reboot without a power cycle. Click OK. The storage system starts the reboot in the specified number of minutes. The reboot takes several minutes. 6. Search for the storage system to reconnect the CMC to the storage system once it has finished rebooting. See “Finding storage systems” (page 17). Powering off the storage system 1. 2. 3. Log in to the storage system. Select Storage System Tasks on the Details tab and select Power Off or Reboot. Select Power Off.
Upgrading the SAN/iQ software on the storage system The CMC enables automated upgrades for the P4000 SAN Solution, including firmware, SAN/iQ upgrades and patches, and CMC upgrades. Use the CMC to check for upgrades and to download and install upgrades and patches. • Check for upgrades for the CMC and storage systems on the SAN. Click the menu bar button, Check for Upgrades, to see if any upgrades are available. • Upgrades are available when the button on the menu bar says Upgrades Available.
• Software Up to Date—System has checked for upgrades, and no upgrades are available. • Upgrades Available—Upgrades are available. Click for more information. • Check for Upgrades—Appears briefly until the system connects to the FTP site to check for upgrades. Persists when the system cannot access the FTP site to check for upgrades. Click to see a message on the Upgrades tab. To check for upgrades online: 1. Click Upgrades Available or Check for Upgrades.
When you upgrade the CMC, the CMC closes and the CMC installer wizard runs. The new version of the CMC opens when the installation is finished. Before upgrading the CMC, verify the following: • You are using Version 9.0 or later of the CMC. • You are using the CMC that you want to upgrade. To upgrade the CMC in an online environment: 1. When you see an available CMC upgrade, click Install next to the CMC icon. The CMC closes, then installs the upgrade. 2.
11. Select the option Only notify me of patches available for my current version of software and click OK. 12. Log in to the management group targeted for the upgrade and click Continue next to the Upgrades Available (Patches). Upgrading storage systems in a management group or available storage systems The system upgrades one storage system at a time in management group, making every effort to keep volumes available. The process includes all patches and upgrades.
The Upgrade Summary window also lists software that must be upgraded before finishing the SAN/iQ upgrade. For any software listed, the current version was downloaded to your download directory. 5. 6. If you are using the listed software, verify that the current versions are installed or install the current versions. After verifying or installing the listed upgrades, select the check box at the bottom of the window, and click Finish.
Figure 7 Availability tab Checking status of dedicated boot devices Some storage systems contain either one or two dedicated boot devices. Dedicated boot devices may be compact flash cards or hard drives. If a storage system has dedicated boot devices, the Boot Devices tab appears in the Storage configuration category. Storage systems that do not have dedicated boot devices will not display the Boot Devices tab. In storage systems with two dedicated boot devices, both devices are active by default.
Table 6 Boot device status (continued) Boot device status Description Unformatted The device has not yet been used in a storage system. It is ready to be activated. Not Recognized The device is not recognized as a boot device. Unsupported The device cannot be used. (For example, the compact flash card is the wrong size or type.) NOTE: When the status of a boot device changes, an event is generated. See “Alarms and events overview” (page 86).
3 Configuring RAID and Managing Disks For each storage system, you can select the RAID configuration and the RAID rebuild options, and monitor the RAID status. You can also review disk information and, for some models, manage individual disks. Getting there 1. 2. In the navigation window, select a storage system and log in if necessary. Open the tree under the storage system and select the Storage category.
Table 7 Descriptions of RAID levels (continued) RAID level Description provides a backup copy of the files and normal system operations are not interrupted. RAID 1+0 – Mirroring and Striping Offers the best combination of data protection and performance. RAID 1+0 or drive mirroring creates fault tolerance by storing duplicate sets of data on a minimum of four hard drives. There must be an even number of drives for RAID 1+0.
RAID devices by RAID type Each RAID type creates different sets of RAID devices. Table 8 (page 34) contains a description of the variety of RAID devices created by the different RAID types as implemented on various storage systems. Table 8 Information in the RAID setup report This item Describes this Device Name The disk sets used in RAID. The number and names of devices varies by storage system and RAID level. Device Type The RAID level of the device.
RAID protects against failure of disks within a storage system, but not against failure of an entire storage system. For example, if network connectivity to the storage system is lost, then data reads and writes to the storage system cannot continue. NOTE: If you plan on using clusters with only a single storage system, use RAID 1 and RAID 10, RAID 5, or RAID 6 to ensure data redundancy within that storage system.
Table 9 Data availability and safety in RAID configurations (continued) Configuration Data safety and availability during disk failure Data availability if entire storage system fails or if network connection to storage system lost Volumes configured with Network RAID-10 or greater on clustered storage systems Yes. However, if any disk in the Yes storage system fails, the entire storage system must be copied from another storage system in the cluster.
3. 4. Change the rebuild settings as desired on the RAID Rebuild Rate Priority window. Click OK. The settings are then ready when a RAID rebuild takes place. Reconfiguring RAID Reconfiguring RAID on a storage system or a VSA destroys any data stored on that storage system. For VSAs, there is no alternate RAID choice, so the only outcome for reconfiguring RAID is to wipe out all data.
When RAID 6 initialization has finished, the CMC becomes available again, and you can add the storage system to the management group. Monitoring RAID status RAID is critical to the operation of the storage system. If RAID has not been configured, the storage system cannot be used. Monitor the RAID status of a storage system to ensure that it remains normal. If the RAID status changes, a CMC event is generated.
Managing disks Use the Disk Setup tab to monitor disk information and perform disk management tasks as listed in Table 10 (page 39). Table 10 Disk management tasks for storage systems Disk setup function Model where available Monitor disk information All Getting there 1. 2. 3. In the navigation window, select a storage system. Select the Storage category in the tree below it. Select the Disk Setup tab.
Table 11 Description of items on the disk report Column Description Disk Corresponds to the physical slot in the storage system.
storage system to the cluster and migrate the volumes and snapshots back to the original cluster. “Changing the cluster—migrating a volume to a different cluster” (page 160). • Perform a cluster swap, replacing the storage systems with the worn out SSD drives with storage systems with new SSD drives. See “Upgrading the storage systems in a cluster using cluster swap” (page 136).
Figure 14 Viewing the disk status of a VSA NOTE: To change the size of the data disk in a VSA, see the HP P4000 VSA Installation and Configuration Guide for instructions about recreating the disk in the VI Client.
Figure 17 Viewing the Disk Setup tab in a HP LeftHand P4300 and HP P4300 G2 Figure 18 Diagram of the drive bays in a HP LeftHand P4300 and HP P4300 G2 Viewing disk status for the HP P4800 and HP P4800 G2 The disks are labeled 1 through 35 in the Disk Setup window( Figure 19 (page 43)), and correspond to the disk drives from top to bottom, left to right, (Figure 20 (page 44)), when you are looking at the front of the HP P4800 and HP P4800 G2.
Figure 20 Diagram of the drive bays in a HP P4800 and HP P4800 G2 Viewing disk status for the HP P4900 G2 The disks are labeled 1 through 8 in the Disk Setup window (Figure 21 (page 44)), and correspond to the disk drives from top to bottom, left to right (Figure 22 (page 44)), when you are looking at the front of the HP P4900 G2.
• Powering the disk on in the CMC (non-hot-swap storage systems) • Rebuilding RAID on the disk or on the storage system (all storage systems) NOTE: Because of the nature of the SSD drive wear management and reporting, it is likely that all the SSD drives in a system will wear out around the same time. Table 12 (page 45) lists disk replacement requirements for specific configurations and storage systems.
Preparing for a disk replacement Use this section to replace a single disk under the following conditions: • You know which disk needs to be replaced through SAN/iQ monitoring. • When viewed in the Disk Setup tab, the Drive Health column shows Marginal (replace as soon as possible) or Faulty (replace right away). • RAID is still on, though it may be degraded and a drive is inactive.
Complete the checklist for replacing a disk in RAID 1, RAID 10, RAID 5, or RAID 6. Then follow the appropriate procedures for the storage system. Replace the disk You may remove and replace a disk from these hot-swap storage systems after checking that the Safe to Remove status indicates “Yes” for the drive to be replaced.
4 Managing the network A physical storage system has two TCP/IP network interfaces (NICs). For each physical storage system you can: • Configure the individual TCP/IP interfaces, including settings for speed and duplex, frame size, and NIC flow control. • Bond NICs to ensure continuous network access or to improve bandwidth. • Set up and manage DNS servers, and a network routing table.
• When configuring a management interface on a P4000 storage system, make sure that interface is on a separate network. Configuring two separate interfaces on the same IP network is not supported and will result in communication problems in the cluster. • When configuring a management interface on a P4000 storage system, be aware that only one interface can be configured with a default gateway, and that interface should be the management interface.
Managing settings on network interfaces Configure or change the settings of the network interfaces in the storage systems. See “Network best practices” (page 48) for more information. Requirements These settings must be configured before creating NIC bonds. TCP status tab Review the status of the TCP interfaces. Change the speed and duplex, frame size, and NIC flow control of an interface. Changing the speed and duplex and NIC flow control can only be performed on interfaces that are not in a bond.
Table 14 Setting storage system speed and duplex settings (continued) Storage system setting speed/duplex Switch setting speed/duplex 100/Full 100/Full 100/Half 100/Half 10/Full 10/Full 10/Half 10/Half NOTE: The VSA does not support changing the speed and duplex settings. Requirements • These settings must be configured before creating NIC bonds. • If you change these settings, you must ensure that both sides of the NIC cable are configured in the same manner.
increasing the frame size requires that routers, switches, and other devices on your network support that frame size. NOTE: Increasing the frame size can cause decreased performance and other network problems if routers, switches, or other devices on your network do not support frame sizes greater than 1500 bytes. If you are unsure about whether your routers and other devices support larger frame sizes, keep the frame size at the default setting.
NOTE: The VSA does not support changing flow control settings. Requirements • These settings must be configured before creating NIC bonds. • All NICs should have (or must have, if they are bonded) the same flow control settings. You can bond NICs with different flow control settings. However, the bond changes both NICs to the same flow control settings, using the settings of the preferred NIC. • Flow control cannot be changed when the port is disabled.
Table 15 Identifying the network interfaces on the storage system Ethernet interfaces Where labeled Label In the TCP/IP Network configuration category in the CMC • eth0, eth1 • TCP/IP tab • Motherboard:Port0, Motherboard:Port1 • TCP Status tab • G4-Motherboard:Port1, G4-Motherboard:Port2 • Motherboard:Port1, Motherboard:Port2 For bonded interfaces: • BondN or Bond0 In the Configuration Interface available through the storage • Intel Gigabit Ethernet system’s serial port • Broadcom Gigabit Ethernet On the
Using DHCP A DHCP server becomes a single point of failure in your system configuration. If the DHCP server goes offline, then IP addresses may be lost. CAUTION: If you use DHCP, be sure to reserve statically assigned IP addresses for all storage systems on the DHCP server. This is required, because management groups use unicast communication. NOTE: If you plan to bond the NICs, you must use a static IP address. To set IP address using DHCP 1. 2. 3. 4. 5. 6.
continues operating normally. To use Link Aggregation Dynamic Mode, your switch must support 802.3ad. CAUTION: Link Aggregation Dynamic Mode requires plugging both NICs into the same switch. This bonding method does not protect against switch failure. • Adaptive Load Balancing (ALB). The logical interface balances data transmissions through both NICs to enhance the functionality of the server and the network. Adaptive Load Balancing automatically incorporates fault tolerance features as well.
Table 16 HP P4000 SAN Solution—possible bonds with 1 GbE and 10 GbE NICs Motherboard:Port1 (1 Motherboard:Port2 (1 eth2 (10 GbE) GbE) GbE) eth3 (10 GbE) SAN/iQ software interface Disabled Disabled Enabled Disabled eth2 Disabled Disabled Disabled Enabled eth3 Disabled Disabled Enabled Enabled eth2 Disabled Disabled Enabled Enabled eth3 Disabled Disabled ALB Bond ALB Bond Bond Disabled Disabled 802.3ad bond 802.
Physical and logical interfaces The two NICs in the storage system are labeled as listed in Table 18 (page 58). If both interfaces are bonded for failover, the logical interface is labeled bond0 and acts as the master interface. As the master interface, bond0 controls and monitors the two slave interfaces which are the physical interfaces.
Table 20 Example Active-Passive failover scenario and corresponding NIC status Example failover scenario NIC status 1. Active-Passive bond0 is created. The active (preferred) interface is Eth0. • Bond0 is the master logical interface. • Eth0 is Active. • Eth1 is connected and is Passive (Ready). 2. Active interface fails. Bond0 detects the failure and Eth1 • Eth0 status becomes Passive (Failed). takes over. • Eth1 status changes to Active. • Eth0 status changes to Active after a 30 second delay. 3.
Figure 26 Active-Passive in a two-switch topology with server failover 1. Servers 2. HP P4000 3. Storage cluster 4. GigE trunk 5. Active path 6. Passive path The two-switch scenario in Figure 26 (page 60) is a basic, yet effective, method for ensuring high availability. If either switch fails, or a cable or NIC on one of the storage systems fails, the Active-Passive bond causes the secondary connection to become active and take over.
Figure 27 Active-Passive failover in a four-switch topology 1. Servers 2. HP P4000 3. Storage cluster 4. GigE trunk 5. Active path 6. Passive path Figure 27 (page 61) illustrates the Active-Passive configuration in a four-switch topology. How link aggregation dynamic mode bonding works Link Aggregation Dynamic Mode allows the storage system to use both interfaces simultaneously for data transfer. Both interfaces have an active status.
Table 22 Link aggregation dynamic mode failover scenario and corresponding NIC status Example failover scenario NIC status 1. Link Aggregation Dynamic Mode bond0 is created. Eth0 and Eth1 are both active. • Bond0 is the master logical interface. • Eth0 is Active. • Eth1 is Active. 2. Eth0 interface fails. Because Link Aggregation Dynamic • Eth0 status becomes Passive (Failed). Mode is configured, Eth1 continues operating. • Eth1 status remains Active. • Eth0 resumes Active status. 3.
Figure 28 Link aggregation dynamic mode in a single-switch topology 1. Servers 2. HP P4000 3. Storage cluster How Adaptive Load Balancing works Adaptive Load Balancing allows the storage system to use both interfaces simultaneously for data transfer. Both interfaces have an active status. If the interface link to one NIC goes offline, the other interface continues operating. Using both NICs also increases network bandwidth.
Table 24 Example Adaptive Load Balancing failover scenario and corresponding NIC status Example failover scenario NIC status 1. Adaptive Load Balancing bond0 is created. • Bond0 is the master logical interface. Motherboard:Port1 and Motherboard:Port2 are both active. • Motherboard:Port1 is Active. • Motherboard:Port2 is Active. 2. Motherboard:Port1 interface fails. Because Adaptive • Motherboard:Port1 status becomes Passive (Failed).
Figure 29 Adaptive Load Balancing in a two-switch topology 1. Servers 2. HP P4000 3. Storage cluster 4. GigE trunk Creating a NIC bond Follow these guidelines when creating NIC bonds: Prerequisites Verify that the speed, duplex, flow control, and frame size are all set properly on both interfaces that are being bonded. These settings cannot be changed on a bonded interface or on either of the supporting interfaces.
• Ensure that the bond has a static IP address for the logical bond interface. The default values for the IP address, subnet mask and default gateway are those of one of the physical interfaces. • Verify on the Communication tab that the SAN/iQ interface is communicating with the bonded interface. CAUTION: To ensure that the bond works correctly, you should configure it as follows: • Create the bond on the storage system before you add it to a management group. • Verify that the bond is created.
NOTE: Because it can take a few minutes for the storage system to re-initialize, the search may fail the first time. If the search fails, wait a minute or two and choose Try Again on the Network Search Failed message. 12. Verify the new bond interface. Figure 31 Viewing a new Active-Passive bond 1. Bonded logical network interface 2. Physical interfaces shown as slaves The bond interface shows as “bond0” and has a static IP address. The two physical NICs now show as slaves in the Mode column. 13.
Figure 32 Verifying interface used for SAN/iQ communication 3. Verify that the SAN/iQ communication port is correct. Viewing the status of a NIC bond You can view the status of the interfaces on the TCP Status tab. Notice that in the Active-Passive bond, one of the NICs is the preferred NIC. In both the Link Aggregation Dynamic Mode bond and the Adaptive Load Balancing bond, neither physical interface is preferred. Figure 33 (page 68) shows the status of interfaces in an Active-Passive bond.
Figure 34 Viewing the status of a link aggregation dynamic mode bond 1. Neither interface is preferred NOTE: If the bonded NIC experiences rapid, sequential Ethernet failures, the CMC may display the storage system as failed (flashing red) and access to data on that storage system fails. However, as soon as the Ethernet connection is reestablished, the storage system and the CMC display the correct information.
Figure 35 Searching for the unbonded storage system on the network 5. Search for the storage system by Host Name or IP Address or Subnet/Mask. NOTE: Because it can take a few minutes for the storage system to re-initialize, the search may fail the first time. If the search fails, wait a minute or two and choose Try Again on the Network Search Failed message. You can also use the Configuration Interface to delete a NIC bond. See “Deleting a NIC bond” (page 237).
Figure 36 Verifying interface used for SAN/iQ communication 5. Verify that the SAN/iQ communication port is correct. Disabling a network interface You can disable the network interfaces on the storage system. • You can only disable top-level interfaces. This includes bonded interfaces and NICs that are not part of bonded interfaces. • To ensure that you always have access to the storage system, do not disable the last interface.
Configuring a disabled interface If one interface is still connected to the storage system but another interface is disconnected, you can reconnect to the second interface using the CMC. See “Configuring the IP address manually” (page 54). If both interfaces to the storage system are disconnected, you must attach a terminal, or PC or laptop to the storage system with a null modem cable and configure at least one interface using the Configuration Interface. See “Configuring a network connection” (page 236).
5. Use the arrows on the Edit DNS Servers window to order the servers. The servers will be accessed in the order they appear in the list. 6. Click OK when you are finished. Adding domain names to the DNS suffixes Add up to six domain names to the DNS suffix list (also known as the look-up zone). The storage system searches the suffixes first and then uses the DNS server to resolve host names. 1. On the DNS tab, click DNS Tasks, and select Edit DNS Suffixes. 2.
6. 7. Click Remove. Click OK to remove the DNS suffix from the list. Setting up routing The Routing tab displays the routing table. You can specify static routes and/or a default route. NOTE: If you specify a default route here, it will not survive a reboot or shutdown of the storage system. To create a route that will survive a storage system reboot or shut down, you must enter a default gateway on the TCP/IP tab. See “Configuring the IP address manually” (page 54).
Configuring storage system communication Use the Communication tab to configure the network interface used by the storage system to communicate with other storage systems on the network and to update the list of managers that the storage system can communicate with. Selecting the interface used by the SAN/iQ software The SAN/iQ software uses one network interface for communication with other storage systems on the network.
Updating the list of manager IP addresses Update the list of manager IP addresses to ensure that a manager running on this storage system is communicating correctly with all managers in the management group. Requirements Each time you update the list of managers, you must reconfigure application servers that use the management group to which this storage system belongs.
5 Setting the date and time The storage systems within management groups use the date and time settings to create a time stamp when data is stored. You set the time zone and the date and time in the management group, and the storage systems inherit those management group settings. • Using network time protocol Configure the storage system to use a time service, either external or internal to your network. • Setting the time zone Set the time zone for the storage system.
NOTE: When using a Windows server as an external time source for an storage system, you must configure W32Time (the Windows Time service) to also use an external time source. The storage system does not recognize the Windows server as an NTP server if W32Time is configured to use an internal hardware clock. 1. Click Time Tasks, and select Add NTP Server. 2. Enter the IP address of the NTP server you want to use. 3. Decide whether you want this NTP server to be designated preferred or not preferred.
The server you added first is the one accessed first when time needs to be established. If this NTP server is not available for some reason, the next NTP server that was added, and is preferred, is used for time serving. To change the order of access for time servers 1. Delete the server whose place in the list you want to change. 2. Add that same server back into the list. It is placed at the bottom of the list, and is the last to be accessed.
6 Administrative users and groups When you create a management group, the SAN/iQ software configures two default administrative groups and one default administrative user. You can add, edit, and delete additional administrative users and groups. All administrative users and groups are managed at the management group level. Getting there In the navigation window, log in to the management group, and select the Administration node.
Adding group membership to a user 1. 2. 3. 4. 5. 6. 7. Log in to the management group, and select the Administration node. Select a user in the Users table. Click Administration Tasks in the tab window, and select Edit User. Click Add in the Member Groups section. Select the groups to which to add the new user. Click OK. Click OK to finish editing the administrative user. Removing group membership from a user 1. 2. 3. 4. 5. 6. Log in to the management group, and select the Administration node.
1. 2. 3. 4. Log in to the management group, and select the Administration node. Click Administration Tasks in the tab window, and select New Group. Enter a Group Name and optional Description. Select the permission level for each function for the group you are creating. See Table 27 (page 82). Table 27 Descriptions of group permissions Management area Activities controlled by this area Change Password User can change other administrative users’ passwords.
• Different levels of access to the storage system, such as read/write • Access to different management capabilities for the storage system, such as creating volumes When you create a group, you also set the management capabilities available to members of a group. The default setting for a new group is Read Only for each category. 3. Click the permission level for each function for the group you are creating. See “Descriptions of group permissions” (page 82) for a description of the permission levels.
7 Monitoring the SAN Monitor the SAN for usage; to ensure that best practices are followed when changes are made, such as adding additional storage systems to clusters; and maintain the overall health of the SAN. Tools for monitoring the SAN include the SAN Status Page, the Configuration Summary and the Best Practice table, the Alarms and Events features, including customized notification methods, and diagnostic tests and log files available for the storage systems.
Figure 39 SAN Status Page 1. Status Page toolbar 2. Content panes Customizing the Status Page content The SAN Status Page includes seven categories of information, displayed in content panes. You can customize the layout of the content panes, as well customize the information displayed in each content pane. • Best Practice Summary—includes all the best practice items that appear on the Configuration Summary node. View the information by management group.
Customizing the Status Page layout Customize the layout of the SAN Status Page to highlight the information most important to you. All customizations are retained when the CMC is closed and restarted. If the Status Page is undocked when the CMC is closed, it will be undocked when the CMC is started again. Drag-and-drop content panes to change their position on the page. The layout is three columns by default. Drag a content pane and drop it on another content pane. The two panes switch places.
Figure 40 Events node and its parts 1. Events toolbar 4. Events list 2. Number of events broken down by severity 5. Alarms list 3. Filter panel NOTE: Except for the P4800 G2 SAN Solution for BladeSystem, alarms and events information is not available for storage systems listed under Available Systems in the CMC, because they are not currently in use on the SAN. Table 28 (page 87) defines the alarms and events columns that appear in the CMC.
Table 28 Alarms and events column descriptions (continued) Column Description Hostname Name of the storage system that posted the event. Source User that generated the event, such as system (software) or a user name. Working with alarms The Alarms window is always visible at the bottom of the CMC and displays all warning and critical events for all management groups you are logged in to. Use the column headings to sort the list of alarms. Double-click an alarm to see more details.
1. 2. 3. In the navigation window, log in to the management group. In the Alarms window, click Alarm Tasks, and select Open Alarms in Window. To close the window, click Close. Exporting alarm data to a .csv file 1. 2. 3. In the navigation window, log in to the management group. In the Alarms window, click Alarm Tasks, and select Export Alarms. In the Filename field, enter a path and file name. If you enter just a file name, the file is stored in the directory that the CMC is installed in. 4. Click OK.
1. 2. 3. In the navigation window, log in to the management group. Select Events in the tree. Click Event Tasks, and select Open Events in Window. The events open in a separate window. Use this window to filter, view details, and export events. 4. To close the window, click Close. Working with events The Events node offers multiple ways to manage and use event data. Filter events to control the information displayed, and export the data for use in a spreadsheet or other analysis tool.
To remove the filters and view all events, click Reset. Saving filter views Save filter views for later use. Custom filters appear in the Filters list in the toolbar below the boldface list of generic filters. Custom filters are available for use with any management group from this installation of the CMC. 1. Click Events Tasks, and select Save Current View as Filter. 2. Give the filter settings a name and click Save. Deleting custom filters 1. 2. 3. 4. Click Event Tasks, and select Delete Filter.
Setting up email notification To set up email notification of events, you must set up the following: • Email server • Email recipients Setting up the email server When you configure a management group, you configure email notification for events by setting up the email server to send events and then adding email recipients to receive notification based on event severity. Change these settings later, if necessary, by using the Email window found under the Events node in the management group tree. 1.
4. 5. 6. 7. 8. Click Add Recipient. In the Address field, enter the email address. Select the following: • Severity of the events • Message type • Language Click OK. Click Apply or Apply and Test. Setting up SNMP The management group can be monitored using an SNMP client. You can also enable SNMP traps to receive system alerts. The Management Information Base (MIB) is read-only and supports SNMP versions 1 and 2c. See “Installing the LeftHand Networks MIB” (page 96) for a list of MIBs.
Configuring access control for SNMP clients Enable the SNMP agent and configure access control for SNMP clients. Enter either a specific IP address and the IP Netmask as None to allow a specific host to access SNMP, or specify the Network Address with its netmask value so that all hosts matching that IP and netmask combination can access SNMP. NOTE: Use the CMC ping feature to verify IP addresses while configuring access control. See “Pinging an IP address” (page 54). Adding an SNMP client 1. 2. 3. 4.
Disabling the SNMP agent Disable the SNMP Agent if no longer using SNMP applications to monitor the network of storage systems. Disabling SNMP 1. 2. 3. 4. In the navigation window, log in to the management group. In the tree, select Events→SNMP. Click SNMP Tasks, and select Edit SNMP General Settings. Select Disable SNMP Agent. Note that the Agent Status field now shows disabled. The SNMP client information remains listed, but cannot be used.
Sending a test trap Send a test trap to verify that your trap recipients are working. 1. In the navigation window, log in to the management group. 2. In the tree, select Events→SNMP. 3. Click SNMP Tasks, and select Send Test Trap. 4. Click OK on the Test SNMP Traps message window. Disabling SNMP traps To disable SNMP traps, you must delete all of the settings in the SNMP Traps window. 1. Remove the Trap Recipient hosts. 2. Delete the Trap Community String. 3. Click OK.
The supported MIBs The following are the supported standard MIBs, though not every function in each MIB is supported. • DISMAN-EVENT-MIB • HOST-RESOURCES-MIB • IF-MIB • IP-FORWARD-MIB • IP-MIB • NET-SNMP-AGENT-MIB • NET-SNMP-EXTEND-MIB • NETWORK-SERVICES-MIB • NOTIFICATION-LOG-MIB • RFC1213-MIB • SNMP-TARGET-MIB • SNMP-VIEW-BASED-ACM-MIB • SNMPv2-MIB • UCD-DLMOD-MIB • UCD-SNMP-MIB Running diagnostic reports Use diagnostics to check the health of the storage system hardware.
List of diagnostic tests Table 29 (page 98) shows a sample of the diagnostic tests that are available for the storage system, including the description of that test and the pass/fail criteria. For each test, lists the following information: • A description of the test • Pass / fail criteria NOTE: Available diagnostic tests depend on the storage system. For VSA, only the Disk Status Test is available.
Figure 41 Opening the hardware information window 1. Link to obtain hardware statistics 2. On the Hardware table, use the link Click to Refresh to obtain the latest hardware statistics. Figure 42 Viewing the hardware information for a storage system Saving a hardware information report 1. 2. 3. Click Hardware Information Tasks and select Save to File to download a text file of the reported statistics. Choose the location and name for the report. Click Save. The report is saved with an .html extension.
Table 30 Selected details of the hardware report This term means this Hostname Hostname of the storage system. Storage system software Full version number for storage system software. Also lists any patches that have been applied to the storage system. IP address IP address of the storage system. Support key Support Key is used by a Technical Support representative to log in to the storage system.
Table 30 Selected details of the hardware report (continued) This term means this Unit number Identifies devices that make up the O/S RAID configuration, including: • Type of storage (BOOT, LOG, SANIQ, DATA) • RAID level (0, 1, 5) • Status (Normal, Rebuilding, Degraded, Off) • Capacity • Rebuild statistics (% complete, time remaining) Controller/cache items Depending on the storage system, this can include: • Information about the RAID controller card and Battery Backup Unit (BBU) • Information about R
To export the summary: 1. From the CMC menu bar, select Tasks→System Summary. 2. Click Export. 3. Select a location for the file, and rename it if desired. 4. Click Export. Configuring a remote log and remote log destination Use remote log files to automatically write log files to a computer other than the storage system. For example, direct the log files for one or more storage systems to a single log server in a remote location. The computer that receives the log files is called the Remote Log Target.
NOTE: After deleting a remote log file from the storage system, remove references to this log file from the syslog configuration on the target computer. Exporting support logs If asked to do so by Customer Support, export support logs for a management group or storage system. The logs are saved as a zip file. 1. Select a management group or storage system in the navigation window. 2.
8 Working with management groups A management group is a collection of one or more storage systems. It is the container within which you cluster storage systems and create volumes for storage. Creating a management group is the first step in creating an IP SAN with the SAN/iQ software. Functions of management groups Management groups serve several purposes: • Management groups are the highest administrative domain for the SAN.
Table 31 Management group requirements (continued) Management group requirement What it means You need the DNS domain name, suffix, and server IP address. See “Using a DNS server” (page 72). Plan email notification You can set up email notification for events for each management group. You must set up the email server to send events. You need the email (SMTP) server IP or host name, server port, a valid email address to use as the sender address. See “Setting up email notification” (page 92).
Add administrative user 1. 2. Click Next to add an administrative user. Enter the administrative user’s name, a description, and a password. The first administrator is always at full administrator level. 3. Click Next to set the time for the management group. Set management group time 1. Select the method by which to set the management group time. • [Recommended] To use an NTP server, know the URL of the server, or its IP address, before you begin.
in the management group. For more information on using the map view tools, see “Using the Map View ” (page 14). Best practice for managers in a management group When creating a management group, the wizard creates an optimal manager configuration for the number of storage systems used to create the group. See Table 32 (page 107) for the default manager configurations.
Table 33 Managers and quorum Number of Managers Number for a quorum Fault tolerance Explanation 1 1 None If the manager fails, no data control takes place. This arrangement is not recommended. 2 2 None Even number of managers not recommended, except in specific configurations. Contact Customer Support for more information. 3 2 High If one manager fails, 2 remain, so there is still a quorum. (Note: 2 managers are not fault tolerant. See above.
Once installed and configured, the Failover Manager operates as a storage system in how you add it to a management group where it serves solely as a quorum tie-breaking manager. Virtual Managers A Virtual Manager is added to a management group, as shown in Figure 44 (page 109), but is not started on a storage system until a failure in the system causes a loss of quorum.
Figure 45 Configuration Summary created when the first management group is configured Configuration guidance As the Configuration Summary reports the numbers of the storage items, it provides warnings about the safe limits for each category, based on performance and scalability. These warnings first alert you that the category is nearing the limits by turning the category orange. When an individual category turns orange, the Configuration Summary category in the navigation window turns orange as well.
of the management group. Exceeding 30 storage systems triggers a warning by turning that line red. As soon as the total number of storage systems reduces below the boundary, the summary bar returns to the previous indicator, either orange or green. Storage systems in the cluster The optimum number of storage systems in a cluster ranges up to 10. If the cluster contains 11 to 16 storage systems, the Configuration Summary appears orange for that line of the management group.
Figure 47 Warning when items in the management group are reaching safe limits 1. Volumes and snapshots are nearing the optimum limit. One cluster is nearing the optimum limit for storage systems. Configuration errors When any item exceeds a recommended maximum, it turns red, and remains red until the number is reduced. See Figure 48 (page 112). Figure 48 Error when some item in the management group has reached its limit 1. Volumes and snapshots have exceeded recommended maximums.
Figure 49 Best Practice Summary for well-configured SAN Expand the management group in the summary to see the individual categories that have recommended best practices. The summary displays the status of each category and identifies any conditions that fall outside the best practice. Click on a row to see details about that item's best practice. Disk level data protection Disk level data protection indicates whether the storage system has an appropriate disk RAID level set.
Volume-level data protection Use a data protection level greater than Network RAID-0 to ensure optimum data availability if a storage system fails. For information about data protection, see “Planning data protection” (page 142). Volume access Use iSCSI load balancing to ensure better performance and better utilization of cluster resources. For more information about iSCSI load balancing, see “iSCSI load balancing” (page 231).
1. 2. 3. In the navigation window, select a management group. Log in by any of the following methods. • Double-click the management group. • Open the Management Group Tasks menu, and select Log in to Management Group. You can also open this menu from a right-click on the management group. • Click any of the Log in to view links on the Details tab. Enter the user name and password, and click Log In.
Stopping managers Under normal circumstances, you stop a manager when you are removing a storage system from a management group. You cannot stop the last manager in a management group. If you stop a manager that compromises fault tolerance, the Best Practice details identify an issue with systems that are running managers. Deleting the management group is the only way to stop the last manager. Implications of stopping managers • Quorum of the storage systems may be decreased.
Table 34 Guide to local bandwidth priority settings Network type Throughput (MB/sec) Throughput rating Minimum 0.25 2 Mbps Ethernet 1.25 10 Mbps Factory Default 4.00 32 Mbps Fast-Ethernet 12.50 100 Mbps Half Gigabit-Ethernet 62.50 500 Mbps Gigabit-Ethernet 128.00 1 Gbps Bonded Gigabit-Ethernet (2) 256.00 2 Gbps Bonded Gigabit-Ethernet (4) 512.00 4 Gbps Set or change local bandwidth priority 1. 2. 3. In the navigation window, select a management group and log in.
Prerequisites • Disconnect any hosts or servers that are accessing volumes in the management group. • Wait for any restriping of volumes or snapshots to complete. • Stop managers on storage systems that you plan to remove from the management group. You may want to add a FOM or start a Virtual Manager on a different storage system to maintain quorum and the best fault tolerance. See “Stopping managers” (page 116). Shut down the management group 1. 2. 3.
the management group is started up, the management group remains in maintenance mode while the repaired or replaced storage system is resynchronizing with the rest of the management group. • After a management group is shut down, a subset of storage systems is powered on. The management group remains in maintenance mode until the remaining storage systems are powered on and rediscovered in the CMC. • For some reason, a storage system comes up, but it is not fully functional.
• Remove the storage system from the cluster. See “Removing a storage system from a cluster” (page 137). • Let any restripe operations finish completely. Remove the storage system 1. 2. 3. 4. Log in to the management group from which you want to remove a storage system. In the navigation window, select the storage system to remove. Click Storage System Tasks on the Details tab, and select Remove from Management Group. Click OK on the confirmation message.
9 Using specialized managers The SAN/iQ software provides two specialized managers that are used in specific situations. The Failover Manager is used in two-system and in Multi-Site SAN configurations to support automatic quorum management in the SAN. A virtual manager is added to a management group but is only started manually on a storage system if needed to maintain or regain quorum. NOTE: You can only use one type of specialized manager in a management group.
• Failover Manager not inside a virtual Windows machine with VMware Server running. • Failover Manager configured to start last after storage systems in the management group in server startup/shutdown list. Using the Failover Manager on Microsoft Hyper-V Server Install the Failover Manager on a Windows Server 2008 from the DVD, or from the DVD .iso image downloaded from the website: http://www.hp.
Using the Failover Manager for VMware Install the Failover Manager from the DVD, or from the DVD .iso image downloaded from the website: http://www.hp.com/go/P4000downloads The installer offers three choices for installing the Failover Manager for VMware. • Failover Manager for ESX—The installer for the Failover Manager for ESX includes a wizard that guides you through configuring the virtual machine on the network and powering on the Failover Manager.
11. Select No, I am done, unless you plan to install another Failover Manager, and click Next. If you want to install another Failover Manager, the wizard repeats the steps, using information you already entered, as appropriate. 12. Finish the installation, reviewing the configuration summary, and click Deploy. When the installer is finished, the Failover Manager is ready to be used in the HP P4000 SAN Solution.
Next, use the Find function in the CMC to discover the Failover Manager, and then add it to a management group. See “Adding a storage system to an existing management group” (page 115) for more information. Installing the Failover Manager for VMware Server or VMware Workstation 1. 2. 3. 4. 5. 6. Begin the installer one of the following ways: • Insert the in the DVD drive. • Double-click the executable that you downloaded to start the installation.
Table 35 Troubleshooting for ESX Server installation (continued) Issue Solution 4. Copy fresh files into the virtual machine folder from the downloaded zip file or distribution media. 5. Open the VI Client, and begin again. You cannot find the Failover Manager with the CMC, and • The CMC displays the IP address of a system if it can cannot recall its IP address. be found. • Open a VI Client session and select the Summary tab for the system you want.
Disaster recovery using a virtual manager The virtual manager functions as an on-demand manager in a disaster-recovery situation. As an on-demand manager, it can be used to regain quorum and maintain access to data. Management group across two sites with shared data Using a virtual manager allows one site to continue operating if the other site fails.
Figure 52 Two-site failure scenarios that are correctly using a virtual manager Configuring a cluster for disaster recovery In addition to using a virtual manager, you must configure your cluster and volumes correctly for disaster recovery. This section describes how to configure your system, including the virtual manager. Best Practice The following example describes configuring a management group with four storage systems in one cluster.
2. Create management group—plan the managers and virtual manager. When you create the management group in the two-site scenario, plan to start two managers per site and add a virtual manager to the management group. You now have five managers for fault tolerance. See “Managers overview” (page 107). 3. Reorder the storage systems in the cluster alternating by site. Create the cluster. Reorder the storage systems to the cluster in alternating order by site, as shown in the bulleted list.
Figure 54 Network RAID-10 volume on two-site cluster Adding a virtual manager If a two-system management group is created without a Failover Manager, a virtual manager is added automatically to the group to ensure fault-tolerance. NOTE: 1. 2. 3. Best practice is to use a Failover Manager for a two-system management group. Select the management group in the navigation window and log in. Click Management Group Tasks on the Details tab, and select Add virtual manager. Click OK to confirm the action.
Starting a virtual manager to regain quorum Only start a virtual manager when it is needed to regain quorum in a management group. “Two-site failure scenarios that are correctly using a virtual manager” (page 128) illustrates the correct way to start a virtual manager when necessary to regain quorum. Two-site scenario, one site becomes unavailable For example, in the two-site disaster recovery model, one of the sites becomes unavailable. On the site that remains up, all managers must be running.
NOTE: If you attempt to start a virtual manager on a storage system that appears to be normal in the CMC, and you receive a message that the storage system is unavailable, start the virtual manager on a different storage system. This situation can occur when quorum is lost, because the CMC may still display the storage system in a normal state, even though the storage system is unavailable.
10 Working with clusters Clusters are groups of storage systems created in a management group. Clusters create a pool of storage from which to create volumes. The volumes seamlessly span the storage systems in the cluster. Expand the capacity of the storage pool by adding storage systems to the cluster.
9. Click Finish. 10. Review the summary information in the next window. Click Close when finished. Adding an iSNS server (option) Add an iSNS server. NOTE: If using an iSNS server, you may not need to add Target Portals in the Microsoft iSCSI Initiator. 1. 2. 3. 4. 5. 6. 7. Select the new cluster in the tree. Click iSCSI Tasks menu, and select Add iSNS Server. Right-click on the cluster or click Cluster Tasks and select Edit Cluster. Select Edit Cluster→Edit iSNS Servers.
3. 4. In the Edit iSNS Servers window, select the VIP to change or delete, or click Add to add a new VIP. Make the necessary changes and click OK to save. Editing iSNS servers from the iSCSI Tasks menu 1. 2. 3. 4. Select the cluster and click the iSCSI tab to bring it to the front. Select the iSNS server IP to change and click iSCSI Tasks. Select the appropriate menu item and make the desired changes. Click OK to save.
Maintaining storage systems in clusters Use the Edit Cluster menu to perform cluster maintenance tasks such as increasing capacity by adding a storage system, upgrading the cluster by swapping in new storage systems, or updating cluster VIPs and iSNS servers when necessary. Adding a new storage system to a cluster Add a new storage system to an existing cluster to expand the storage for that cluster.
5. 6. Select the storage systems to swap into the cluster and click OK. Click OK to complete the swap. The swap operation may take some time, depending upon the number of storage systems swapped and the amount of data being restriped. Reordering storage systems in a cluster Reorder the systems in a cluster to control the stripe patterns, especially in a multi-site cluster. 1. Select the cluster in the navigation window. 2. Click Cluster Tasks, and select Edit Cluster→Reorder Storage Systems. 3.
Auto Performance Protection is indicated by two unique statuses reported on the Details tab. You will also receive event notifications on these statuses. For information about setting up event notification, see “Configuring events” (page 89). • Storage System Overloaded. The Overloaded status indicates that operations to the storage system are completing too slowly. During the overloaded state, volume availability is maintained while the storage system is quarantined in the cluster.
If status is Storage System Inoperable Reboot the storage system and see if it returns to Normal, when it comes back up. If these statuses recur This may be an indication that the underlying hardware problem still exists. Repairing a storage system Repairing a storage system allows you to replace a failed disk in a storage system that contains volumes configured for data protection levels other than Network RAID-0, and trigger only one resync of the data, rather than a complete restripe.
3. From the Repair Storage System window, select the item that describes the problem to solve. Click More for more detail about each selection. • Repair a disk problem If the storage system has a bad disk, be sure to read “Replacing a disk” (page 44) before beginning the process. • Storage system problem Select this choice if you have verified that the storage system must be removed from the management group to fix the problem.
11 Provisioning storage The SAN/iQ software uses volumes, including SmartClone volumes, and snapshots to provision storage to application servers and to back up data for recovery or other uses. Before you create volumes or configure schedules to snapshot a volume, plan the configuration you want for the volumes and snapshots. Planning your storage configuration requires understanding how the capacity of the SAN is affected by the RAID level of the storage systems and the features of the SAN/iQ software.
Full provisioning Full provisioning reserves the same amount of space on the SAN as is presented to application servers. Full provisioning ensures that the application server will not fail a write. When a fully provisioned volume approaches capacity, you receive a warning that the disk is nearly full. Thin provisioning Thin provisioning reserves less space on the SAN than is presented to application servers. The SAN/iQ software allocates space as needed when data is written to the volume.
Data protection level Seven data protection levels are available, depending upon the number of available storage systems in the cluster. Table 38 Setting a data protection level for a volume With this number of Select any of these data protection levels available storage systems in cluster For this number of copies 1 • Network RAID-0 (None) • One copy of data in the cluster. 2 • Network RAID-0 (None) • One copy of data in the cluster.
How data protection levels work The system calculates the actual amount of storage resources needed for all data protection levels. When you choose Network RAID-10, Network RAID-10+1, or Network RAID-10+2, data is striped and mirrored across either two, three, or four adjacent storage systems in the cluster. When you choose Network RAID-5 or Network RAID-6, the layout of the data stripe, including parity, depends on both the Network RAID mode and cluster size.
Best applications for Network RAID-10+1 are those that require data availability even if two storage systems in a cluster become unavailable. Figure 59 (page 145) illustrates the write patterns on a cluster with four storage systems configured for Network RAID-10+1. Figure 59 Write patterns in Network RAID-10+1 (3-Way Mirror) Network RAID-10+2 (4-Way Mirror) Network RAID-10+2 data is striped and mirrored across four or more storage systems.
Figure 61 (page 146) illustrates the write patterns on a cluster with four storage systems configured for Network RAID-5. Figure 61 Write patterns and parity in Network RAID-5 (Single Parity) 1. Parity for data blocks A, B, C 2. Parity for data blocks D, E, F 3. Parity for data blocks G, H, I 4. Parity for data blocks J, K, L Network RAID-6 (Dual Parity) Network RAID-6 divides the data into stripes and adds parity.
Figure 62 Write patterns and parity in Network RAID-6 (Dual Parity) 1. P1 is parity for data blocks A, B, C, D 2. P2 is parity for data blocks E, F, G, H 3. P3 is parity for data blocks I, J, K, L 4. P4 is parity for data blocks M, N, O, P Provisioning snapshots Snapshots provide a copy of a volume for use with backup and other applications. You create snapshots from a volume on the cluster. Snapshots are always thin provisioned.
Plan how you intend to use snapshots, and the schedule and retention policy for schedules to snapshot a volume. Snapshots record changes in data on the volume, so calculating the rate of changed data in the client applications is important for planning schedules to snapshot a volume. NOTE: Volume size, provisioning, and using snapshots should be planned together. If you intend to use snapshots, review “Using snapshots” (page 162).
Figure 63 Cluster tab view Cluster use summary The Use Summary window presents information about the storage space available in the cluster. Figure 64 Reviewing the Use Summary tab In the Use Summary window, the Storage Space section lists the space available on the storage systems in the cluster. Saved space lists the space saved in the cluster by using thin provisioning and the SmartClone feature.
Table 39 Information on the Use Summary tab (continued) Category Description snapshots are created, or as thinly provisioned volumes grow. Saved Space Thin Provisioning The space saved by thin provisioning volumes. This space is calculated by the system. SmartClone Feature Space saved by using SmartClone volumes is calculated using the amount of data in the clone point and any snapshots below the clone point. Only as data is added to an individual SmartClone volume does it consume space on the SAN.
Table 40 Information on the Volume Use tab (continued) Category Description to see the space saved number decrease as data on the volume increases. • Full provisioning allocates the full amount of space for the size of the volume. Reclaimable space is the amount of space that you can get back if this fully provisioned volume is changed to thinly provisioned. Consumed Space Amount of space used by actual data volumes or snapshots.
Table 41 Information on the System Use tab Category Description Name Host name of the storage system. Raw space Total amount of disk capacity on the storage system. Note: Storage systems with greater capacity will only operate to the capacity of the lowest capacity storage system in the cluster. RAID configuration RAID level configured on the storage system. Usable space Space available for storage after RAID has been configured.
However, the file system does not inform the block device underneath (the SAN/iQ volume) that there is freed-up space. In fact, no mechanism exists to transmit that information. There is no SCSI command which says “Block 198646 can be safely forgotten.” At the block device level, there are only reads and writes. So, to ensure that our iSCSI block devices work correctly with file systems, any time a block is written to, that block is forever marked as allocated.
Changing configuration characteristics to manage space Options for managing space on the cluster include • Changing snapshot retention—retaining fewer snapshots requires less space • Changing schedules to snapshot a volume—taking snapshots less frequently requires less space • Deleting volumes or moving them to a different cluster NOTE: Deleting files on a file system does not free up space on the SAN volume. For more information, see “Block systems and file systems” (page 152).
12 Using volumes A volume is a logical entity that is made up of storage on one or more storage systems. It can be used as raw data storage or it can be formatted with a file system and used by a host or file server. Create volumes on clusters that contain one or more storage systems.
Table 43 Characteristics for new volumes Volume characteristic Configurable for Primary or Remote Volume What it means Volume Name Both The name of the volume that is displayed in the CMC. A volume name is from 1 to 127 characters and is case sensitive. The volume name cannot be changed. You can enable and customize a default naming convention for volumes. See “Setting naming conventions” (page 15) for more information. Description Both [Optional] A description of the volume.
Table 43 Characteristics for new volumes (continued) Volume characteristic Configurable for Primary or Remote Volume What it means Type Both • Primary volumes are used for data storage. • Remote volumes are used for configuring Remote Copy for business continuance, backup and recovery, or data mining/migration. Default value = Primary Provisioning Primary • Fully provisioned volumes are the same size on the SAN as the size presented to the application server.
NOTE: The system automatically factors data protection levels into the settings. For example, if you create a fully provisioned 500 GB volume and the data protection level is Network RAID-10 (2–Way Mirror), the system automatically allocates 1000 GB for the volume. To set advanced characteristics for a volume, continue on the Advanced tab of the New Volume window. Configuring advanced volume settings [optional] Set additional characteristics for a volume in the Advanced tab in the New Volume window.
Table 44 Requirements for changing volume characteristics Item Requirements for Changing Description Must be from 1 to 127 characters. Server Server must have already been created in the management group.
Changing the volume description 1. 2. In the Description field, edit the description. Click OK when you are finished. Changing the cluster—migrating a volume to a different cluster Editing the volume to change the cluster causes the volume to migrate to the target cluster. The SAN/iQ software allows the online migration of volumes between storage clusters in the same management group. This migration can be used to manage both storage capacity and performance between clusters.
Deleting a volume Delete a volume to remove that volume’s data from the storage system and make that space available. Deleting a volume also deletes all the snapshots underneath that volume, except for clone points and shared snapshots. For more information, see “Clone point” (page 185) and “Shared snapshot ” (page 187). CAUTION: Deleting a volume permanently removes that volume’s data from the storage system.
13 Using snapshots Snapshots are a copy of a volume for use with backup and other applications. Types of snapshots Snapshots are one of the following types: • Regular or point-in-time —Snapshot that is taken at a specific point in time. However, an application writing to that volume may not be quiesced. Thus, data may be in flight or cached and the actual data on the volume may not be consistent with the application's view of the data.
would run weekly and retain 5 copies. A third schedule would run monthly and keep 4 copies. • File-level restore without tape or backup software • Source volumes for data mining, test and development, and other data use. Best Practice—Use SmartClone volumes. See “SmartClone volumes” (page 178) . Planning snapshots When planning to use snapshots, consider their purpose and size.
Table 46 Snapshot characteristics (continued) Snapshot parameter What it means vCenter Server is installed. See the HP P4000 Application Aware Snapshot Manager Deployment Guide for more information about the controlling server IP address. Prerequisites for application-managed snapshots Creating an application-managed snapshot using the SAN/iQ software is the same as creating any other snapshot. However, you must select the Application-Managed Snapshot option in the New Snapshot window.
Creating an application-managed snapshot, with or without volume sets, requires the use of the Application Aware Snapshot Manager. The application-managed snapshot option quiesces Windows and VMware applications on the server before creating the snapshot. For more information, see “Prerequisites for application-managed snapshots” (page 164). If the Application Aware Snapshot Manager is not installed, the SAN/iQ software creates a regular, point-in-time snapshot.
Editing a snapshot You can edit both the description of a snapshot and its server assignment. The description must be from 0 to 127 characters. 1. Log in to the management group that contains the snapshot that you want to edit. 2. In the navigation window, select the snapshot. 3. Click Snapshot Tasks on the Details tab, and select Edit Snapshot. 4. Change the description as necessary. 5. Change the server assignment as necessary. 6. Click OK when you are finished.
Requirements for snapshot schedules Table 49 Characteristics for creating a schedule to snapshot a volume Item Description and requirements Name The name of the schedule that is displayed in the CMC. A scheduled snapshot name must be from 1 to 127 characters and is case sensitive. Snapshots created by a schedule have a default naming convention enabled when the CMC is installed. You can change or disable this naming convention.
Creating a schedule to snapshot a volume 1. 2. 3. 4. 5. In the navigation window, select volume for which you want to create a schedule for snapshots. Click Volume Tasks on the Details tab and select New Schedule to Snapshot a Volume. Enter a name for the schedule. (Optional) Enter a snapshot description. Click Edit to specify a start date and time. The Date and Time Configuration window opens. Use this window to set the date and time for the first snapshot created by this schedule. 6. 7. 8.
6. (Optional) If displayed, click Verify Volume Associations to see if the volume set included in the snapshot set is up to date. The Verify Volume Associations window opens, showing the volumes currently associated with the schedule. Any volumes that have been added to or removed from the volume set are reflected. Click Close to return to the Edit Schedule to a Snapshot Volume window. The updated list of volumes is populated in the Volume Set field.
Figure 68 Delete multiple snapshots from the volumes and snapshots node Scripting snapshots Application-based scripting allows automatic snapshots of a volume. For detailed information, see “Working with scripting” (page 196) and the HP CLIQ - The SAN/iQ Command-Line Interface User Guide, for information about the SAN/iQ software command-line interface. Mounting a snapshot A snapshot is a copy of a volume.
Making a Windows application-managed snapshot available If you do any of the following using a Windows application-managed snapshot, you must use diskpart.
Making a Windows application-managed snapshot available on a server in a Microsoft cluster Use this procedure to make an application-managed snapshot available on servers that are in a Microsoft cluster. NOTE: 1. 2. 3. 4. 5. 6. We recommend contacting Customer Support before performing this procedure. Disconnect the iSCSI sessions. Do one of the following (based on what you need to do with the application-managed snapshot): • Convert temporary space. • Create a SmartClone.
Managing snapshot temporary space You can either delete the temporary space to free up space on the cluster, or, if you need data that may have been written to the temporary space, convert that temporary space to a SmartClone volume. Convert the temporary space to access data Convert the snapshot temporary space if you have written data to a mounted snapshot and you need to permanently save or access that data.
between the original volume and the new SmartClone volume. For detailed information about SmartClone volumes, see “What are SmartClone volumes?” (page 178). • Use Remote Copy to copy the newer snapshots that you want to keep, before performing the rollback. See the HP P4000 Remote Copy User Guide for more information about copying data. Best practices for rolling back a volume • Stop any applications that are accessing the volume, and log off all related iSCSI sessions.
If the snapshot is not part of a snapshot set If the snapshot is part of a snapshot set • Click Roll Back Volume. See “Continue with standard roll back” (page 175). NOTE: This will leave some snapshot sets incomplete. This is not recommended. • Click Cancel. See “Cancel the rollback operation” (page 176). Continue with standard roll back The following steps result with the original volume, with its original name, returned to the state of the rolled back snapshot.
Figure 69 New volume with shared clone point 1. Original volume 2. New SmartClone volume from snapshot 3. Shared clone point 5. If you created the SmartClone from an application-managed snapshot, use diskpart.exe to change the resulting volume's attributes. For more information, see “Making a Windows application-managed snapshot available” (page 171).
Restrictions on deleting snapshots You cannot delete a snapshot when the snapshot is: • A clone point. • In the process of being deleted or being copied to a remote management group. • The primary snapshot that is copied using Remote Copy, and you are not logged into the remote management group that it is copied to. CAUTION: Typically, you do not want to delete individual snapshots that are part of a snapshot set.
14 SmartClone volumes SmartClone are space-efficient copies of existing volumes or snapshots. They appear as multiple volumes that share a common snapshot, called a clone point. They share this snapshot data on the SAN. SmartClone volumes can be used to duplicate configurations or environments for widespread use, quickly and without consuming disk space for duplicated data. Use the SmartClone process to create up to 25 volumes in a single operation.
Table 50 Terms used for SmartClone features (continued) Term Definition (page 179), the snapshots Volume_1_SS_1 and Volume_1_SS_2 are shared snapshots. Map view Tab that displays the relationships between clone points and SmartClone volumes. See the map view in Figure 82 (page 191) and Figure 83 (page 192). In Figure 70 (page 179) you can see on the left a regular volume with three snapshots and on the right, a regular volume with one SmartClone volume, one clone point, and two shared snapshots.
Safely use production data for test, development, and data mining Use SmartClone volumes to safely work with your production environment in a test and development environment, before going live with new applications or upgrades to current applications. Or, clone copies of your production data for data mining and analysis. Test and development Using the SmartClone process, you can instantly clone copies of your production LUNs and mount them in another environment.
Naming convention for SmartClone volumes A well-planned naming convention helps when you have many SmartClone volumes. Plan the naming ahead of time, since you cannot change volume or snapshot names after they have been created. You can design a custom naming convention when you create SmartClone volumes. Naming and multiple identical disks in a server Mounting multiple identical disks to servers typically requires that servers write new disk signatures to them.
Table 51 Characteristics for new SmartClone volumes (continued) SmartClone volume characteristic What it means more information, see “Assigning server connections access to volumes” (page 203). Permission Type of access to the volume: Read, Read/Write, None Naming SmartClone volumes Because you may create dozens or even hundreds of SmartClone volumes, you need to plan the naming convention for them.
Figure 73 Rename SmartClone volume from base name 1. Rename SmartClone volume in list Shared versus individual characteristics Characteristics for SmartClone volumes are the same as for regular volumes. However, certain characteristics are shared among all the SmartClone volumes and snapshots created from a common clone point.
Figure 74 Programming cluster with SmartClone volumes, clone point, and the source volume 1. Source volume 2. Clone point 3. SmartCLone volumes (5) In this example, you edit the SmartClone volume, and on the Advanced tab you change the cluster to SysAdm. The confirmation message lists all the volumes and snapshots that will change clusters as a result of changing the edited volume.
Figure 76 SysAdm cluster now has the SmartClone volumes, clone point, and the source volume Table 52 (page 185) shows the shared and individual characteristics of SmartClone volumes. Note that if you change the cluster or the data protection level of one SmartClone volume, the cluster and data protection level of all the related volumes and snapshots will change.
clone point. That is, you can delete all but one of the SmartClone volumes, and then you can delete the clone point. Figure 77 Navigation window with clone point 1. Original volume 2. Clone point 3. SmartClone volume In Figure 77 (page 186), the original volume is “C#.” • Creating a SmartClone volume of C# first creates a snapshot, C#_SCsnap. • After the snapshot is created, you create at least one SmartClone volume, C#class_1.
Figure 78 Clone point appears under each SmartClone volume 1. Clone point appears multiple times. Note that it is exactly the same in each spot NOTE: Remember that a clone point only takes up space on the SAN once. Shared snapshot Shared snapshots occur when a clone point is created from a newer snapshot that has older snapshots below it in the tree. They are designated in the navigation window with the icon shown here.
Figure 79 Navigation window with shared snapshots 1. Original volume 2. Clone point 3. Shared snapshots In Figure 79 (page 188), the original volume is C#. Three snapshots were created from C#: • C#_snap1 • C#_snap2 • C#_SCsnap Then a SmartClone volume was created from the latest snapshot, C#_SCsnap. That volume has a base name of C#_class. The older two snapshots, C#_snap1 and C#_snap2, become shared snapshots, because the SmartClone volume depends on the shared data in both those snapshots.
Creating SmartClone volumes You create SmartClone volumes from existing volumes or snapshots. When you create a SmartClone volume from another volume, you first take a snapshot of the original volume. When you create a SmartClone volume from a snapshot, you do not take another snapshot. To create a SmartClone volume When you create SmartClone volumes, you either set the characteristics for the entire group or set them individually. Figure 80 Setting characteristics for SmartClone volumes 1.
5. 6. 7. 8. Next you select the following characteristics: • Base name for the SmartClone volumes • Type of provisioning • Server you want connected to the volumes, and • Appropriate permission. In the Quantity field, select the number of SmartClone volumes you want to create. Click Update Table to populate the table with the number of SmartClone volumes you selected. If you want to modify any individual characteristic, do it in the list before you click OK to create the SmartClone volumes.
Figure 82 Viewing SmartClone volumes and snapshots as a tree in the Map View Using views The default view is the tree layout, displayed in Figure 82 (page 191). The tree layout is the most effective view for smaller, more complex hierarchies with multiple clone points, such as clones of clones, or shared snapshots. You may also display the Map view in the organic layout.
Figure 83 Viewing the organic layout of SmartClone volumes and related snapshots in the Map View Viewing clone points, volumes, and snapshots The navigation window view of SmartClone volumes, clone points, and snapshots includes highlighting that shows the relationship between related items. For example, in Figure 84 (page 193), the clone point is selected in the tree. The clone point supports the SmartClone volumes, so it is displayed under those volumes.
Figure 84 Highlighting all related clone points in navigation window 1. Selected clone point 2. Clone point repeated under SmartClone volumes Editing SmartClone volumes Use the Edit Volume window to change the characteristics of a SmartClone volume. Table 55 Requirements for changing SmartClone volume characteristics Item Shared or Individual Requirements for Changing Description Individual May be up to 127 characters. Size Individual Sets available space on cluster.
Table 55 Requirements for changing SmartClone volume characteristics (continued) Item Shared or Individual Requirements for Changing Type Individual Determines whether the volume is primary or remote. Provisioning Individual Determines whether the volume is fully provisioned or thinly provisioned. To edit the SmartClone volumes 1. 2. In the navigation window, select the SmartClone volume for which you want to make changes. Click Volume Tasks, and select Edit Volume.
Deleting multiple SmartClone volumes Delete multiple SmartClone volumes in a single operation from the Volume and Snapshots node of the cluster. First you must stop any application servers that are using the volumes, and log off any iSCSI sessions. 1. Select the Volumes and Snapshots node to display the list of SmartClone volumes in the cluster. Figure 86 List of SmartClone volumes in cluster 2. 3. Use Shift+Click to select the SmartClone volumes to delete. Right-click, and select Delete Volumes.
15 Working with scripting Scripting in the SAN/iQ software through release 7.0 was accomplished by the java.commandline.CommandLine scripting tool. In SAN/iQ software release 8.0, the java.commandline.CommandLine scripting tool was replaced by the HP P4000 Command-Line Interface. The CLI takes advantage of the new SAN/iQ API that provides comprehensive coverage of SAN/iQ software functionality to support scripting, integration and automation. The java.commandline.
16 Controlling server access to volumes Application servers (servers), also called clients or hosts, access storage volumes on the SAN using the iSCSI protocol. You set up each server that needs to connect to volumes in a management group in the CMC. We refer to this setup as a “server connection.” Server connections can be single servers or server clusters. Use server clusters in the CMC to easily assign volumes to the clustered application servers that are accessing the SAN storage volumes.
Figure 87 Server assignments in the navigation window and the Volumes and Snapshots tab 1. Servers from a version 7.0 management group in the navigation window 2. Volumes and Snapshots tab shows the assigned volume that the server can access Adding server connections to management groups Add each server connection that needs access to a volume to the management group containing the volume.
Table 57 Characteristics for servers (continued) Item Description and requirements Initiator Node Name The name, or iqn string, of the iSCSI initiator. Open the iSCSI initiator and look for the string there. You can copy the string and paste it into the field. Enable load balancing Configures iSCSI load balancing for the server connections. CHAP not required Enables initiators to log on to a volume without authenticating their identity.
Editing server connections You can edit the following fields for a server connection: • Description • Controlling Server IP Address • Load balancing • CHAP options CAUTION: If you change the load balancing or CHAP options, you must log off and log back on to the target in the iSCSI initiator for the changes to take effect. You can also delete a server connection from the management group. For more information, see “Deleting server connections” (page 200).
Creating a server cluster 1. 2. 3. 4. 5. 6. In the navigation window, select the Servers category. Right-click and select New Server Cluster. Enter a name and description (optional) for the server cluster. Do one of the following: • Click Add Server and select the server from the list of available servers that opens. • Click New Server and follow the instructions found in “Adding server connections to management groups” (page 198).
Figure 89 Completed server cluster and the assigned volumes 1. Green solid line indicates active connection. The two-way arrows indicate the volume permission levels are read-write. Black dotted line indicates an inactive iSCSI session. 2. Volumes have active iSCSI sessions, indicated by the line to the left of the volume icon Server cluster map view After you create a server cluster and connect volumes, use the Map View tab for viewing the relationships between systems, volumes and servers.
Figure 90 Servers and volumes retain connections after server cluster is deleted 1. Each volume remains connected to each server after the server cluster is deleted To delete a server cluster and remove connections: 1. In the navigation window, select the Servers category and select the server cluster to delete. 2. Right-click on the server cluster and select Delete Server Cluster. 3. Select a server to change associations. 4. Right-click and select Assign and Unassign Volumes and Snapshots. 5.
Table 59 Server connection permission levels Type of Access Allows This No access Prevents the server from accessing the volume or snapshot. Read access Restricts the server to read-only access to the data on the volume or snapshot. Read/write access Allows the server read and write permissions to the volume. NOTE: Microsoft Windows requires read/write access to volumes. Assigning server connections from a volume You can assign one or more server connections to any volume or snapshot.
1. 2. 3. 4. In the navigation window, right-click the volume whose server connection assignments you want to edit. Select Assign and Unassign Servers. Change the settings as needed. Click OK. Editing server assignments from a server connection You 1. 2. 3. 4. can edit the assignment of one or more volumes or snapshots to any server connection. In the navigation window, right-click the server connection you want to edit. Select Assign and Unassign Volumes and Snapshots. Change the settings as needed.
17 Monitoring performance The Performance Monitor provides performance statistics for iSCSI and storage system I/Os to help you and HP support and engineering staff understand the load that the SAN is servicing. The Performance Monitor presents real-time performance data in both tabular and graphical form as an integrated feature in the CMC. The CMC can also log the data for short periods of time (hours or days) to get a longer view of activity.
Generally, the Performance Monitor can help you determine: • Current SAN activities • Workload characterization • Fault isolation Current SAN activities example This example shows that the Denver cluster is handling an average of more than 747 IOPS with an average throughput of more than 6 million bytes per second and an average queue depth of 31.76.
Figure 93 Example showing fault isolation What can I learn about my volumes? If you have questions such as these about your volumes, the Performance Monitor can help: • Which volumes are accessed the most? • What is the load being generated on a specific volume? The Performance Monitor can let you see the following: • Most active volumes • Activity generated by a specific server Most active volumes examples This example shows two volumes (DB1 and Log1) and compares their total IOPS.
Figure 95 Example showing throughput of two volumes Activity generated by a specific server example This example shows the total IOPS and throughput generated by the server (ExchServer-1) on two volumes.
Figure 97 Example showing network utilization of three storage systems Load comparison of two clusters example This example illustrates the total IOPS, throughput, and queue depth of two different clusters (Denver and Boulder), letting you compare the usage of those clusters. You can also monitor one cluster in a separate window while doing other tasks in the CMC.
Figure 99 Example comparing two volumes Accessing and understanding the Performance Monitor window The Performance Monitor is available as a tree system below each cluster. To display the Performance Monitor window: 1. In the navigation window, log in to the management group. 2. Select the Performance Monitor system for the cluster you want. The Performance Monitor window opens. By default, it displays the cluster total IOPS, cluster total throughput, and cluster total queue depth.
For more information about the performance monitor window, see the following: • “Performance Monitor toolbar” (page 212) • “Performance monitor graph” (page 212) • “Performance monitor table” (page 213) Performance Monitor toolbar The toolbar lets you change some settings and export data. Figure 101 Performance Monitor toolbar Button or Status Definition 1. Performance Monitor status • Normal—Performance monitoring for the cluster is OK.
Figure 102 Performance Monitor graph The graph shows the last 100 data samples and updates the samples based on the sample interval setting. The vertical axis uses a scale of 0 to 100. Graph data is automatically adjusted to fit the scale. For example, if a statistic value was larger than 100, say 4,000.0, the system would scale it down to 40.0 using a scaling factor of 0.01. If the statistic value is smaller than 10.0, for example 7.5, the system would scale it up to 75 using a scaling factor of 10.
Table 60 Performance Monitor table columns (continued) Column Definition Units Unit of measure for the statistic. Value Current sample value for the statistic. Minimum Lowest recorded sample value of the last 100 samples. Maximum Highest recorded sample value of the last 100 samples. Average Average of the last 100 recorded sample values. Scale Scaling factor used to fit the data on the graph’s 0 to 100 scale. Only the line on the graph is scaled; the values in the table are not scaled.
Table 61 Performance Monitor statistics Statistic Definition IOPS Reads Cluster Volume or Snapshot NSM Average read requests X per second for the sample interval. X X IOPS Writes Average write requests X per second for the sample interval. X X IOPS Total Average read+write X requests per second for the sample interval. X X Throughput Reads Average read bytes per second for the sample interval. X X X Throughput Writes Average write bytes per second for the sample interval.
Table 61 Performance Monitor statistics (continued) Statistic Definition Cluster Volume or Snapshot NSM system for the sample interval. Memory Utilization Percent of total memory used on this storage system for the sample interval. - X Network Utilization Percent of bidirectional network capacity used on this network interface on this storage system for the sample interval. - X Network Bytes Read Bytes read from the network for the sample interval.
Access size The size of a read or write operation. As this size increases, throughput usually increases because a disk access consists of a seek and a data transfer. With more data to transfer, the relative cost of the seek decreases. Some applications allow tuning the size of read and write buffers, but there are practical limits to this. Access pattern Disk accesses can be sequential or random.
3. Click . Figure 105 Add Statistics window 4. From the Select Object list, select the cluster, volumes, and storage systems you want to monitor. Use the CTRL key to select multiple objects from the list. 5. From the Select Statistics options, select the option you want. • All Statistics—Adds all available statistics for each selected object. • Selected Statistics from List—Lets you select the statistics you want from the list below.
3. Right-click a row in the table and select View Statistic Details. The Statistic Details window opens, with all of the information for the selected statistic that is in the table, plus the statistic definition. 4. Click Close.
Pausing and restarting monitoring If you are currently monitoring one or more statistics, you can pause the monitoring and restart it. For example, you may want to pause monitoring during planned maintenance windows or production downtime. 1. From the Performance Monitor window, click to pause the monitoring. All data remain as they were when you paused. 2. To restart the monitoring, click . Data updates when the next sample interval elapses. The graph will have a gap in the time.
1. From the Performance Monitor window, right-click the statistic in the table that you want to highlight, and select Highlight. The line turns white. 2. To remove the highlight, right-click the statistic, and select Remove Highlight. Changing the scaling factor The vertical axis uses a scale of 0 to 100. Graph data is automatically adjusted to fit the scale. For example, if a statistic value was larger than 100, say 4,000.0, the system would scale it down to 40.0 using a scaling factor of 0.01.
9. Click Add Statistics. The statistics you selected are listed in the Added Statistics list. 10. Click OK. The File Size field displays an estimated file size, based on the sample interval, duration, and selected statistics. 11. When the export information is set the way you want it, click OK to start the export. The export progress appears in the Performance Monitor window, based on the duration and elapsed time. To pause the export, click To stop the export, click , then click to resume the export.
18 Registering advanced features Advanced features expand the capabilities of the SAN/iQ software and are enabled by licensing the storage systems through the HP License Key Delivery Service website, using the license entitlement certificate that is packaged with each storage system. However, you can use the advanced features immediately by agreeing to enter an evaluation period when you begin using the SAN/iQ software for clustered storage.
Identifying licensing status You can check the status of licensing on individual advanced features by the icons displayed. The violation icon appears throughout the evaluation period. Figure 106 Identifying the license status for advanced features Backing out of Remote Copy evaluation If you decide not to use Remote Copy and you have not obtained license keys by the end of the evaluation period, you must delete any remote volumes and snapshots you have configured.
5. 6. Read the text, and select the box to enable the use of scripts during a license evaluation period. Click OK. Turn off scripting evaluation Turn off the scripting evaluation period when you take either one of these actions: • You obtain license keys for the feature you were evaluating. • You complete the evaluation and decide not to license any advanced features. Turning off the scripting evaluation ensures that no scripts continue to push the evaluation clock.
Submitting storage system feature keys 1. 2. 3. 4. 5. 6. In the navigation window, select the storage system from the Available Systems pool for which you want to register advanced features. Select the Feature Registration tab. Select the Feature Key. Right-click, and select Copy. Use Ctrl+V to paste the feature key into a text editing program, such as Notepad. Register and generate the license key at the Webware website: https://webware.hp.
The Registration tab displays the following information: • The license status of all the advanced features, including the progress of the evaluation period and which advanced features are in use and not licensed • Version information about software components of the operating system • Customer information (optional) Submitting storage system feature keys Submit the feature keys of all the storage systems in the management group. 1.
NOTE: Record the host name or IP address of the storage system along with the feature key. This record will make it easier to add the license key to the correct storage system when you receive it. Entering license keys When you receive the license keys, add them to the storage systems in the Feature Registration window. 1. In the navigation window, select the management group. 2. Select the Registration tab. 3. Click Registration Tasks, and select Feature Registration from the menu. 4.
Make a customer information file for each management group in your SAN. • Create or edit your customer profile. • Save the customer profile to a computer that is not part of your SAN. Editing your customer information file Occasionally, you may want to change some of the information in your customer profile. For example, if your company moves, or contact information changes. 1. In the navigation window, select a management group. 2. Click the Registration tab. 3.
19 iSCSI and the HP P4000 SAN Solution The SAN/iQ software uses the iSCSI protocol to let servers access volumes. For fault tolerance and improved performance, use a VIP and iSCSI load balancing when configuring server access to volumes.
iSCSI load balancing Use iSCSI load balancing to improve iSCSI performance and scalability by distributing iSCSI sessions for different volumes evenly across storage systems in a cluster. iSCSI load balancing uses iSCSI Login-Redirect. Only initiators that support Login-Redirect should be used. When using VIP and load balancing, one iSCSI session acts as the gateway session. All I/O goes through this iSCSI session.
Figure 110 Differentiating types of CHAP CHAP is optional. However, if you configure 1-way or 2-way CHAP, you must remember to configure both the server and the iSCSI initiator with the appropriate characteristics. Table 65 (page 232) lists the requirements for configuring CHAP.
NOTE: The initiator node name and secrets set in the SAN/iQ CMC must match what you enter in the server’s iSCSI initiator exactly. Sample iSCSI configurations Figure 111 (page 233) illustrates the configuration for a single host authentication with CHAP not required with Microsoft iSCSI. Figure 111 Viewing the MS iSCSI initiator to copy the initiator node name Figure 112 (page 233) illustrates the configuration for a single host authentication with 1-way CHAP required.
Figure 113 (page 234) illustrates the configuration for a single host authentication with 2-way CHAP required. Figure 113 Adding an initiator secret for 2-way CHAP (shown in the MS iSCSI initiator) CAUTION: Without the use of shared storage access (host clustering or clustered file system) technology, allowing more than one iSCSI application server to connect to a volume concurrently without cluster-aware applications and/or file systems in read/write mode could result in data corruption.
20 Using the Configuration Interface The Configuration Interface is the command line interface that uses a direct connection with the storage system. You may need to access the Configuration Interface if all network connections to the storage system are disabled. Use the Configuration Interface to perform the following tasks.
$ xterm 3. In the xterm window, start minicom as follows: $ minicom -c on -l NSM Opening the Configuration Interface from the terminal emulation session 1. 2. 3. Press Enter when the terminal emulation session is established. Enter start, and press Enter at the log in prompt. When the session is connected to the storage system, the Configuration Interface window opens.
Table 68 Identifying Ethernet interfaces on the storage system (continued) Ethernet Interfaces Where labeled What the label says Motherboard:Port1, Motherboard:Port2 Configuration Interface Intel Gigabit Ethernet or Broadcom Gigabit Ethernet Label on the back of the storage system Eth0, Eth1, or a graphical symbol similar to the following: or Once you have established a connection to the storage system using a terminal emulation program, you can configure an interface connection using the Configuratio
TCP speed and duplex. You can change the speed and duplex of an interface. If you change these settings, you must ensure that both sides of the NIC cable are configured in the same manner. For example, if the storage system is set for Auto/Auto, the switch must be set the same. For more information about TCP speed and duplex settings, see “Managing settings on network interfaces” (page 50). Frame size. The frame size specifies the size of data packets that are transferred over the network.
21 Support and other resources Contacting HP For worldwide technical support information, see the HP support website: http://www.hp.
New and changed information in this edition The following additions and changes have been made for this edition: • The following information has been updated: ◦ P4000 SAN Solution and user documentation have been rebranded ◦ New SAN Status Home page is added to the CMC, providing a graphic view for monitoring the SAN ◦ New Best Practice analyzers have been added ◦ Management Groups, Clusters and Volume wizard has been updated with elements to ensure a highly available SAN configuration ◦ New snaps
A Replacing hardware This chapter describes the disk replacement procedures for cases in which you do not know which disk to replace and/or you must rebuild RAID on the entire storage system. For example, if RAID has gone off unexpectedly, you need HP Support to help determine the cause, and if it is a disk failure, to identify which disk must be replaced. It also describes how to identify and replace the RAID controller in the P4900 G2 storage system.
1. 2. Log in to the management group. Select the storage system in the navigation window, and review the Details tab information. If the Storage System Status shows Manager Normal, and the Management Group Manager shows Normal, then a manager is running and needs to be stopped. To stop a manager: 1. Right-click the storage system in the navigation window and select Stop Manager.
Rebuilding data The following steps take you through the steps to first reconfigure RAID on the storage system and then to rebuild data on the storage system after it is added to the management group and cluster. Reconfigure RAID 1. 2. Select the Storage category, and select the RAID Setup tab. Click RAID Setup Tasks, and select Reconfigure RAID. The RAID Status changes from Off to Normal. NOTE: If RAID reconfigure reports an error, reboot the storage system, and try reconfiguring the RAID again.
Restarting a manager NOTE: Before proceeding, verify that the storage system has finished initializing and is completely added to the management group. If necessary, ensure that after the repair you have the appropriate configuration of managers. If there was a manager running on the storage system before you began the repair process, you may start a manager on the repaired storage system as necessary to finish with the correct number of managers in the management group.
Controlling server access Use the Local Bandwidth Priority setting to control server access to data during the rebuild process: • When the data is being rebuilt, the servers that are accessing the data on the volumes might experience slowness. Reduce the Local Bandwidth Priority to half of its current value for immediate results. • Alternatively, if server access performance is not a concern, raise the Local Bandwidth Priority to increase the data rebuild speed. To change local bandwidth priority: 1.
Figure 116 Storage server LEDs 1. Front UID/LED switch 2. System health LED 3. NIC 1 activity LED 4. NIC 2 activity LED 5.
Figure 117 Card 1 location Figure 118 Card 2 location A cache module is attached to each RAID controller and each cache module is connected to a battery. The unit is called a backup battery with cache (BBWC). BBWC 1 connects to Card 1 and BBWC 2 connects to Card 2. Removing the RAID controller 1. 2. 3. Power off the storage system: a. Use the CMC to power off the system controller as described in “Powering off the storage system” (page 24). b. Manually power off the disk enclosure.
4. Remove the top cover (Figure 119 (page 248)): a. Loosen the screw on the top cover with the T-10 wrench. b. Press the latch on the top cover. c. Slide the cover toward the rear of the server and then lift the top cover to remove it from the chassis. Lift the top cover away from the chassis. Figure 119 Removing the cover 5. Locate and remove the PCI cage: a.
6. The cache module is attached to the RAID controller and must be removed before removing the RAID controller. Each cache module is connected to a battery; observe the BBWC status LED (4, Figure 121 (page 249)) on both batteries before removing a cache module: • If the LED is flashing every two seconds, data is trapped in the cache. Reassemble the unit, restore system power, and repeat this procedure. • If the LED is not lit, continue with the next step of removing the RAID controller.
Figure 123 Removing the cache module 9. Remove the RAID controller from its slot. Installing the RAID controller IMPORTANT: The replacement RAID controller contains a new cache module. You must remove the cache module on the replacement controller board and attach the existing cache module to the replacement controller board and reconnect the cache module to the battery cable. 1. 2. Slide the RAID controller into the slot, aligning the controller with its matching connector.
Figure 125 Installing Card 2 3. Reinstall the PCI cage (Figure 126 (page 251)): a. Align the PCI cage assembly to the system board expansion slot, and then press it down to ensure full connection to the system board. b. Tighten the thumbscrews to secure the PCI cage assembly to the system board and secure the screw on the rear panel of the chassis. Figure 126 Reinstalling the PCI cage 4. 5. 6. 7. Place the cover back on the unit.
B Third-party licenses The software distributed to you by HP includes certain software packages indicated to be subject to one of the following open source software licenses: GNU General Public License (“GPL”), the GNU Lesser General Public License (“LGPL”), or the BSD License (each, an “OSS Package”).
C SANiQ TCP and UDP Port Usage A P4000 SAN uses a list of well-known TCP/UDP ports to operate, see Table 71 (page 253). Management applications in SAN/iQ software Management applications include the Centralized Management Console and the scripting interface. These applications all use the ports described as management applications in the description column.
Table 71 TCP/UDP ports used for normal SAN operations with SAN/iQ (continued) IP Protocol Port(s) Name Description storage systems. Used for management and control. TCP 13840, 13850, 13851 SAN/iQ Internal Outgoing from management applications. Incoming to storage systems. Used for management and control. TCP 13846, 13850, 13946, 13847 SAN/iQ Support Trace Ports for SAN/iQ Support only. Not required for normal day-to-day operations.
Table 71 TCP/UDP ports used for normal SAN operations with SAN/iQ (continued) IP Protocol Port(s) Name Description information using a web interface. TCP 2301, 2302, 2381, 2382 HP System Insight Manager Used by the HP System Insight Manager tool to display configuration information of HP platforms like DL380, DL320S using a web interface. SSH 16022 TCP 11120 SAN/iQ Upgrade Incoming to storage systems. Used for upgrading the storage systems. Not required for normal day-to-day operations.
Glossary The following glossary provides definitions of terms used in the SAN/iQ software and the HP P4000 SAN Solution. acting primary volume The remote volume, when it assumes the role of the primary volume in a failover scenario. Active-Passive A type of network bonding which, in the event of a NIC failure, causes the logical interface to use another NIC in the bond until the preferred NIC resumes operation. At that point, data transfer resumes on the preferred NIC.
DSM Device Specific Module. DSM for MPIO The HP P4000 DSM for MPIO vendor-specific DSM that interfaces with the Microsoft MPIO framework. failback After failover, the process by which you restore the primary volume and turn the acting primary back into a remote volume. failover The process by which the user transfers operation of the application server over to the remote volume. This can be a manual operation, a scripted operation, or VMware enabled.
Multi-Site cluster A cluster of storage that spans multiple sites (up to three). A Multi-Site cluster must meet at least one of the following conditions: • Contain storage systems that reside in two or more sites • Contain storage systems that span subnets • Contain multiple VIPs. The cluster can have a single site, and the multiple VIPs make it a multi-site cluster. network RAID Synchronous replication, mirroring or parity protection on a volume-by-volume basis.
RAID status Condition of RAID on the storage system: • Normal - RAID is synchronized and running. No action is required. • Rebuild - A new disk has been inserted in a drive bay and RAID is currently rebuilding. No action is required. • Degraded - RAID is not functioning properly. Either a disk needs to be replaced or a replacement disk has been inserted in a drive. • Off - Data cannot be stored on the storage system. The storage system is offline and flashes red in the network window.
shared snapshot Shared snapshots occur when a clone point is created from a newer snapshot that has older snapshots below it in the tree. All the volumes created from the clone point will display these older snapshots that they share, as well as the clone point. site A user-designated location in which storage systems are installed. Multi-Site SAN configurations have multiple sites with storage systems in each site, and each site has its own subnet.
volume set Two or more volumes used by an application. For example, you may set up Exchange to use two volumes to support a StorageGroup: one for mailbox data and one for logs. Those two volumes make a volume set. volume size The size of the virtual device communicated to the operating system and the applications. VSS Provider HP P4000 VSS Provider is the hardware provider that supports the Volume Shadow Copy Service on the HP P4000 SAN Solution.
Index Symbols 1000BASE T interface, 53 802.
B backing out of Remote Copy evaluation, 224 of scripting evaluation, 225 bandwidth, changing local settings, 117 best practice configuring cluster for disaster recovery, 128 frame size, 52 link aggregation dynamic mode, 56 network, 48 NIC bonds, 65 recommended numbers for management group storage items, 110 setting volume size, 142 speed and duplex settings, 51 using snapshots for protection against data deletion, 162 Best Practice Summary cluster-level data protection, 113 disk level data protection, 113
changing for volumes, 160 cluster-level data protection Best Practice Summary, 113 clustered application servers changing volume associations after deleting server cluster, 203 creating, 201 deleting cluster, 202 editing, 202 clustering application servers, 197, 200 requirements for, 200 clustering managers, 104 clustering servers, 197, 200 clusters adding, 133 adding storage system, 136 capacity, 133 comparing the load of two, 210, 216 data protection levels, 143 deleting, 140 editing, 134 map view, 134 ov
using RAID for, 34 data transmission, 52 date setting with NTP, 77 setting without NTP, 79 date and time for scheduled snapshot, 168 decreasing volume size, 160 defaults font size in CMC, 15 language displayed in CMC, 15 naming conventions in CMC, 15 restoring for the Performance Monitor, 219 definition RAID configurations, 32 SmartClone volumes, 178 degraded RAID, 38 deleting administrative groups, 83 administrative users, 81 an administrative group, 83 clone point, 194 clusters, 140 custom event filters,
and static IP addresses, 72 editing IP or domain name, 73 removing, 73 using, 72 document related information, 240 documentation HP website, 240 Domain Name Server see DNS Server domain names adding to DNS suffixes, 73 editing in DNS suffixes list, 73 removing from DNS suffixes list, 73 downloading upgrades, 25 DSM for MPIO, 230 how to see if using, 234 tips for using to access volumes from servers, 234 when using two NICs, 205 duplex, configuring, 50 Dynamic Host Configuration Protocol see DHCP E editing
full permissions, 82 G gateway session for VIP with load balancing, 231 Getting Started Launch Pad, 16 ghost storage system, 139 removing after data is rebuilt, 245 Gigabit Ethernet, 53 see also GBe glossary for SAN/iQ software and HP P4000 SAN, 256 SmartClone volumes, 178 graphical legend Centralized Management Console, 13 group name editing, 82 groups administrative, 81 administrative default groups, 81 deleting administrative, 83 H hardware diagnostics, 97 list of diagnostic tests, 98 tab window, 97 ha
setting in preferences, 15 locating a storage system in a rack, 22 log files saving for technical support, 101 log in to a storage system in a management group, 21 to management group, 13, 114 to storage systems in Available Systems pool, 13 log out of management group, 115 logging on to volumes in iSCSI, 205 logs exporting support, 103 M maintenance mode changing to normal mode, 119 management group in, 118 management group caution for logging into, 13 management group time refreshing, 77 management group
network interface bonds, 55 active-passive, 57 adaptive load balancing, 63 best practices, 65 communication after deleting, 70 configuring, 65 creating, 66 deleting, 69, 237 determining if use would improve performance, 209 link aggregation dynamic mode, 61 physical and logical interface, 58 requirements, 56 requirements for Adaptive Load Balancing, 63 setting flow control, and, 53 settings, 57 status of, 68 using 1 GbE and 10 GbE interfaces, 56 verifying, 67 VSA, 48 with 10 GbE interfaces, 56 network inter
hiding, 220 hiding a line, 220 showing, 220 Performance Monitor window accessing, 211 graph, 212 parts defined, 211 saving to an image file, 222 table, 213 toolbar, 212 permissions administrative group, 82 effect of levels, 204 full, 82 read modify, 82 read only, 82 ping IP address, 54 planning SmartClone volumes, 180 snapshots, 147, 163 volume size, 141 volumes, 141, 155 planning capacity full provisioning method, 142 thin provisioning method, 142 point-in-time snapshots defined, 162 pool of storage, 104 p
when not running manager, 241 rebuild volume data, 244 rebuilding RAID, 47 rate for RAID, 36 reclaimable space in volumes, 150 reconfiguring RAID, 37 recreate the RAID array, 243 redundant array of independent disks see RAID registering add-on applications, 225 registering features Feature Registration tab, 225, 227 for a storage system, 29 related documentation, 240 Remote Copy backing out of evaluation, 224 evaluating, 223 registering, 225 remote volumes, 155 remote destinations setting up for events, 89
SAN/iQ upgrading, 25, 28 saved space in cluster, 149 saved space in volumes, 150 saving diagnostic reports, 97 log files for technical support, 101 management group configuration information, 117 scaling factor changing, 221 scheduled snapshots requirements for, 166 scripting evaluation, 224 backing out of, 225 turning off, 225 turning on, 224 searching for storage systems, 13, 19 security administrative, 104 of storage resources, 104 server access SmartClone volumes, 181 server cluster, 197, 200 change vol
planning size, 163 point-in-time, 162 read/write and deleting temporary space, 173 requirements for editing, 166 restrictions on deleting, 177 rolling back a volume from, 173 shared, 187 size, 148 temporary space for read/write snapshots, 173 understanding schedules for volume sets, 167 using, 162 versus backups, 162 SNMP access control, 94 clients, adding, 94 configuring access control for SNMP clients, 94 disabling agents, 95 disabling traps, 96 editing trap recipients, 95 enabling agents, 93 enabling age
system requirements for Failover Manager on ESX Server, 123 systems finding on network, 13, 19 T TCP frame size, 51 speed and duplex, 50 status, 50 status tab, 50 TCP/IP tab, 53 technical support HP, 239 saving log files for, 101 service locator website, 240 temporary space deleting, 173 for read/write snapshots, 173 making application-managed snapshot available after converting, 172 thresholds changing for a snapshot, 166 requirements for changing in snapshots, 166 time editing NTP server, 78 NTP servers,
deleting application-managed snapshots for, 161, 177 volume size best practice for setting, 142 volume use window, 150 volume-level data protection Best Practice Summary, 114 volumes adding, 157 application server requirements for rolling back, 174 Assign Volume and Snapshot wizard, 18 assigning to servers, 203, 204 changing clusters, 160 changing data protection levels, 160 changing descriptions, 159 changing size, 160 comparing the load of two, 210 consumed space, 150 controlling server access to, 197 cre