Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
General and some specific questions in regards to initial the Distributor database setup and failover
1. How are you guys configuring the distributor database? Is this just the standard local distributor setup? and not the one outlined here:
2. the SQL Agent jobs that gets created when running the configured Qlik task for the first time are only created on the primary. Do we need to manually create those tasks on the other servers in the AG as well?
3. When failing to another server in the AG, we get an alert this alert, The process could not execute 'sp_replcmds' on 'Server01', but replication continues to work. The alert goes away when i manually stop the agent from running. Is this the proper way to handle this?
Dear @KhanhMan ,
Thank you for reaching out to us in the Qlik Community portal.
If our response has been helpful, please consider clicking "Accept as Solution". This will assist other users in easily finding the answer.
Regards,
Deepak
A little bit of research, testing, and help from you folks here has yielded this setup for me and it looks to be working.
1. I had to configure a remote distributor using this link: https://www.sqlshack.com/configure-sql-server-replication-for-a-database-in-sql-server-always-on-ava...
2. The steps outlined in the link that was important for me was to make sure to execute the stored procedure for the redirected publisher. This was the step that took care of the error, The process could not execute 'sp_replcmds', when I would fail over to other replicas in the AG.
3. After a few rounds of testing, I purposefully turned of the original primary and found that the Qlik replicate task would fail saying that it couldn't connect to the server that was powered off to mimic an automatic failover.
4. Searching the forums, I found that enabling this setting of the Internal Parameters "AlwaysOnSharedSynchedBackupIsEnabled" fixed that error.
Hi @KhanhMan
I am not able to directly address your questions but wanted to ensure you had reviewed our documentation on the subject as it may address at least some of your questions: Working with Microsoft SQL Server AlwaysOn Availability Groups | Qlik Replicate Help
Hope this helps,
Dana
Hello team,
If our response has been helpful, please consider clicking "Accept as Solution". This will assist other users in easily finding the answer.
Regards,
Sushil Kumar
I believe I figured it out but still running a few tests with failovers. There's still a couple of items that aren 't working like it should and once those are figured out, I will outline the steps taken to get it to all work happily together.
I've commented on this before. Prior to SQL 2016, distribution db was not supported in AG. That changed in 2016 and is described in the link you provided. However, you can still configure a remote distribution db as described here:
In that case, the jobs you mention in item 2 will be created on the remote distributor. The distributor is not protected by AG failover. But when we've encountered any issues with the distributor agent jobs, Qlik Replicate will still function since it's basically looking for articles to search for in the TLOG. You may however, encounter issues with the TLOG not clearing the log_reuse_wait flag.
Dear @KhanhMan ,
Thank you for reaching out to us in the Qlik Community portal.
If our response has been helpful, please consider clicking "Accept as Solution". This will assist other users in easily finding the answer.
Regards,
Deepak
A little bit of research, testing, and help from you folks here has yielded this setup for me and it looks to be working.
1. I had to configure a remote distributor using this link: https://www.sqlshack.com/configure-sql-server-replication-for-a-database-in-sql-server-always-on-ava...
2. The steps outlined in the link that was important for me was to make sure to execute the stored procedure for the redirected publisher. This was the step that took care of the error, The process could not execute 'sp_replcmds', when I would fail over to other replicas in the AG.
3. After a few rounds of testing, I purposefully turned of the original primary and found that the Qlik replicate task would fail saying that it couldn't connect to the server that was powered off to mimic an automatic failover.
4. Searching the forums, I found that enabling this setting of the Internal Parameters "AlwaysOnSharedSynchedBackupIsEnabled" fixed that error.
Hi Deepak,
Does the same apply when using a SQL Server target? e.g putting the IP of the listener?
Regards,
Mohammed