HP StorageWorks Storage Mirroring Recover User's Guide (T5437-96008, November 2009)

180 of 739
1. By default, Active Monitoring Enabled is selected. Active Monitoring allows you
to enable or disable failover for the selected source/target pair.
2. By default, Manual Intervention Required is selected. Manual intervention allows
you to control when failover occurs. When a failure occurs, a prompt appears and
waits for you to initiate the failover process manually.
Disable Manual Intervention Required only if you want failover to occur
immediately when a failure condition is met.
3. In the Method to monitor for Failover field, select the ping method to use when
monitoring source IP addresses.
● Network Access (ICMP)—Storage Mirroring Recover failover uses ICMP
pings to determine if the source server is online. If a network device, such as
a firewall or router, between the source and target is blocking ICMP traffic,
failover monitors cannot be created or used.
● Replication Service (UDP)—The Storage Mirroring service on the target
server sends a ping-like UDP request to the source Storage Mirroring
service, which replies immediately to confirm it is running. This method is
useful when ICMP is blocked on routers between the source and target.
● Application Monitoring—If you select the Application Monitoring option,
the fields in Application Monitoring Settings area will be enabled.
Note:
In order to use Application Monitoring, the monitored servers must be running
Storage Mirroring Recover version 5.2 or later.
Application monitoring is not available when dealing with a cluster.
4. Configure how long to wait after a source failure to initiate failover. The amount of
time before failover begins is calculated by multiplying the Monitor Interval by the
Failure Count. For example, if the Monitor Interval is set to 5 seconds and the
Failure Count setting is 5, a failover condition will be identified after 25 seconds of
missed source activity.
Note:
In a cluster environment, make sure to include a cushion to account for the
time it takes to failover between nodes. Setting values too low will cause the
Application Manager to assume the source cluster has failed and will
indicate that a failover condition has been met.
a. In the Monitor Interval field, select how often the monitor should check the
source machine availability (Default = 5 seconds, or 60 seconds if
Application Monitoring is selected).
b. In the Failure Count field, select how many monitor replies can be missed
before assuming the source machine has failed (Default = 5 missed replies,
or 3 replies if Application Monitoring is selected).