Unsolved
This post is more than 5 years old
2 Posts
0
899
June 18th, 2013 17:00
Networker, Cluster Enabler and Windows Cluster Failovers
Hello,
In our SQL environment we are using Cluster Enabler to automatically fail-over SAN drives between our SQL servers when we fail-over our SQL instances between the active and inactive node. Effectively when we perform an windows cluster fail-over of SQL, the N,O and P drives disappear from the server that was active before the fail-over and appear on the server which is to be active after the fail-over.
One of the issues we have is that when a cluster fail-over occurs, the Networker backup job for the drives then needs to be manually stopped on the old active node and started on new active node. If we just leave the backup jobs running on both nodes, Networker generates a lot of error noise in our monitoring systems for the inactive node because the drives to be backed up are not present.
Is there a way around this. Is there a way to set up Networker so that it will not attempt the backup jobs on a node when the drives are not present?
Polska1422
2 Intern
•
128 Posts
0
June 19th, 2013 06:00
Hi,
You'll need to use the virtual server name for the SQL server instead of node names for the backup job.
Regards
Claudio
ble1
6 Operator
•
14.4K Posts
•
56.2K Points
0
June 19th, 2013 15:00
Yes, you should configure it to run in cluster - see documentation. That means you will have cluster client too (apart from physical nodes you already have) and configure cluster owned backups to go through virtual client instead.
Jagil
2 Posts
0
June 19th, 2013 19:00
Thanks for the replies, but sorry, I forgot a vital piece of information.
Our servers have 3 separate interfaces and networks they connect to. Production, Backup and Management. The virtual client exists on the production network only and not the backup network, so we cannot access the active server over the backup network and security rules dictate we can't make the connection into the production network.
ble1
6 Operator
•
14.4K Posts
•
56.2K Points
0
June 20th, 2013 03:00
Then change it in configuration. What you can do is initial (metadata) contact over production and send data back over backup (assuming physical box has it). If not, then you can't have setup as per configuration guide and have to live with what you have. Assuming physical box has backup interface, I see no issue why not add backup interface to virtual one.