3.3
Table Of Contents
- Avid iNEWS Administration Guide
- Contents
- Using This Guide
- 1 - Introduction
- 2 - Connect Services
- 3 - Database Security
- 4 - Database Management
- 5 - Backing Up the iNEWS System
- 6 - Disconnects
- 7 - Troubleshooting
- A - Command References
- Programs Invoked by iNEWS
- Commands Used by Avid Personnel Only
- Linux Commands Used in iNEWS
- Console Control Commands
- Console Server Commands
- broadcast
- configure
- connect
- ctraits
- dbclean
- dbclose
- dbdev
- dbdump
- dbfree
- dblines
- dboriginal
- dbpurge (Superuser conditional)
- dbrestore
- dbserver
- dbsort
- dbtraits
- dbvisit
- dictionary
- diskclear (Superuser only)
- diskcopy
- doc
- ed
- enter
- force (Superuser only)
- grpcheck
- gtraits (Superuser only)
- help
- hogs
- idiff
- list
- list B
- list C
- list c
- list d
- list g
- list p
- list q
- list s
- list sq
- list u
- logout
- makemontab
- makeshift (Super user only)
- maketab (Superuser only)
- msgclean
- offline
- online
- otod
- reconnect
- remove
- rename (Superuser only)
- reorder
- restart
- searchtape
- send
- shutdown
- sitedump (Superuser only)
- siterestore (Superuser only)
- startup
- status
- stop
- su
- unbusy
- utraits (Super user only)
- version
- wholockedit
- Job List Commands
- Dialog Commands
- B - System Files
- C - Standard Dictionaries
- Using Dictionaries to Define Messages and Commands
- Customizing Dictionaries
- Utility Messages Dictionary (/site/dict/messages)
- DBServer Program Messages
- Disconnect Program Messages
- Category and Keyword Check Program Messages
- Keyboard Check Program Messages
- Keyboard Check Program Messages for Macros
- Grpcheck Messages
- Wire Program Messages
- Mail Server Messages
- Validation (Action) Server
- Seek Server Messages
- Last Login Messages
- Print Server Messages
- dbtraits Messages
- Save Error (Workstation) Messages
- Queues Dictionary (/site/dict/queues)
- Words Dictionary (/site/dict/words)
- Keyboard Macros Dictionary (/site/dict/keymacros)
- Case-shifting Dictionary (/site/dict/shift)
- MCS Dictionary (/site/dict/mcs)
- Job List Command Dictionary (/site/dict/joblist)
- D Messages Dictionary (/site/dict/dmessages)
- S Messages Dictionary (/site/dict/smessages)
- D - Environment Variables
- E - Managing Traits at the Console
- F - The Line Editor, ed
- Index
Disconnect Recovery
89
In normal dual-server operation, half the devices and sessions are configured on one server
and the other half are configured on the other server. The most important thing to do after a
disconnect is to reconfigure the survivor so that it knows it must be responsible for all
devices and sessions. You can then restart any network devices that were running on the
failed server.
n
The steps are covered in more detail in the next section, “Procedures” on page 89.
The disconnect recovery steps are as follows:
1. Reboot the failed server.
2. Clear the database on the failed server.
3. Reconnect the failed server.
4. Copy the database to the failed server.
5. Log out and stop all servers and resources.
6. Reconfigure the system.
7. Start up all servers and resources.
Procedures
This section contains an example of recovery from a disconnect of a dual AB system. The
steps might vary when the system is a triple system configuration. Triple system
configuration customers might need to contact Avid for more specific instructions.
In the example, it is assumed that server A was chosen as the survivor and server B was
designated the failed server:
•NRCS-A–survivor
• NRCS-B–failed server
The failed server is also referred to as the revived server after it is reconnected to the system.
n
If you reverse the roles of the servers during your disconnect and make B the survivor and
server A the failed server, remember to adjust which server you issue the commands on
appropriately. The steps shown in the following example assume server A is the survivor;
reverse A and B throughout the process if server B is chosen as your survivor.
It is assumed (for this example) that when the servers disconnected, server A became a
single system (system is A, master is A) and server B also became a single system (system is
B, master is B). In this situation, users logged in on A can continue working and saving
stories. Users logged in on B can also continue working and saving stories, but since the
servers are not mirroring, anything saved to server A’s database is not copied over to server
B’s database, and vice versa.