We've been having a lot of database performance issues lately, and I've been trying to see if I can figure out why. We don't have a DBA (I'm a software developer) so I'm sort of just winging it, and much of what I find online reads like a foreign language to me.
We've been restarted SQL Server every morning because that's the only way it's operational during the working day. I noticed that every morning around 5 am we are started to get this message every two minutes in the log:
FlushCache: cleaned up 11848 bufs with 7432 writes in 97168 ms (avoided 8139 new dirty bufs) for db 9:0
last target outstanding: 4, avgWriteLatency 32
average throughput: 0.72 MB/sec, I/O saturation: 11635, context switches 18849
The numbers differ each time of course, but it's the same message over and over in that pattern until I restart the server. I'm not sure how to interpret this, I've been trying to Google about it and all I've gathered is that it means there might be something wrong with I/O and that something is taking longer than it's supposed to. We recently switched to using SSD so I didn't think it should be a write problem.
Could anyone shed some light on this?