Comments
Sort by recent activity
Peter, thanks for the advice and however I can't afford for it to break again. This is a critical DR log shipping process from production systems. If I have further problems after switching to USESIMPLECOPY I'll give this a go but if USESIUMPLECOPY works okay I'll leave it alone to settle down for a while.
Having said that does this version have any changes to the log copying process or does it just contain extra debugging code? / comments
Peter, thanks for the advice and however I can't afford for it to break again. This is a critical DR log shipping process from production systems. If I have further problems after switching to USES...
Update on my approach.
The log copy queue is still getting stuck even with copying to a local disk. Multiple files get stuck in "Copying" and I have a couple that are stuck on "Pending" newer files are going through.
I'm going to try switching to USESIMPLECOPY on the backup command that recommended for local copies anyway? / comments
Update on my approach.
The log copy queue is still getting stuck even with copying to a local disk. Multiple files get stuck in "Copying" and I have a couple that are stuck on "Pending" newer files...
Another issue which might mean I actually leave it this way is that if it has long network copy operations it's very difficult to stop the service if you need to. I actually needed to stop the service and I tried and eventually it failed, even with a long time out period and it actually caused the database cluster to fail over at an inconvenient time. I've since set this service to not cause the cluster to fail over and no restart but had to learn it the hard way (this is not ideal either). With the robocopy method the copying doesn't effect the backup service and so this feels a lot safer to me as I'm not sure what would happen if a fault caused the cluster to failover and the sqlbackup service had to wait minutes to shutdown. / comments
Another issue which might mean I actually leave it this way is that if it has long network copy operations it's very difficult to stop the service if you need to. I actually needed to stop the serv...
The lower the thread count the worse it seemed to perfom. At low thread counts it seemed to get a lot of the logs stuck in pending that would never start. I bumped up the thread count to ensure that they all at least tried to start before the next log backup otherwise the early log backups seemed to never get a chance. Our log files can vary a lot in size and as such sometimes it doesn't catch up copying until a couple of 15 minute backup cycles.
To get over this for the mean time I've set the copy to go to a local folder to consolidate the logs from the backup into one location which I then move out to the remote server through a robocopy move operation. / comments
The lower the thread count the worse it seemed to perfom. At low thread counts it seemed to get a lot of the logs stuck in pending that would never start. I bumped up the thread count to ensure tha...