Table of Contents
Advertisement
Tools for special tasks
7.5.2 Example of how this works
Procedure when a system fails:
System A fails, System B continues to operate:
1. Activate additional B controller with the IP address of A.
2. Switch disks.
3. Provide resources (from A) on the active system (B). These resources must be
identical.
4. Transfer access permissions for data from A (from the regular backup) to data on B.
5. Activate the interface of the additional A controller (on B).
6. Activate servername <A> on B.
7. Servername <A> is operational for the user with data from A.
Procedure when the failed system is operational again:
System A is operational again (system is idle, the switchover is canceled again by the
administrator):
1. Delete existing connections by user of A on B.
2. Deactivate servername <A> (on B).
3. Deactivate interface of additional A controller on B.
4. Deactivate additional controller on B with the IP address of A.
5. Export and save access permissions for A data on B.
6. Delete shares (from A) on B if necessary (or leave the empty directory as is).
7. Switch disks again.
8. Remove unused entries from the ACL database (on B) if necessary.
9. Transfer access permissions for A data (on A) (from the backup on B).
10. Servername <A> is operational again for the user with the data from A.
The above example only shows the most important steps involved in this type of procedure.
The OBSERVE scripts for high availability provide a more detailed description of the mutual
system dependencies when a system fails.
Sample OBSERVE scripts are also supplied with the OBSERVE product. Should you require
further information on operating OBSERVE, please contact you local support center.
Product Manual
OBSERVE support
U7613-J-Z815-6-76
Advertisement
Table of Contents
Troubleshooting
