I just upgraded from M2 all the way to 3.0.1. It went quite smooth and I was pleasantly surprised by how well it worked. Everything was retained from configuration to mailboxes. I've logged in and made sure I'm still getting mail. Nice!
Unfortunately I have one fairly significant problem which appears to have come about from the upgrade:
I noticed that the CPU was pegged, thats strange, I checked and it was mysqld. Okay, its gotta be rebuilding something from the upgrade. I checked again about an hour later and its still pegged. I thought it strange, so I dug around a little more and found that the file logger_myslow.log has grown significantly, about 4 MB in the last hour. Sure enough the process thats killing the cpu is tied to the perl monitor process.
Looking at the log I'm a little baffled. This perl process appears to poll the database every 5 seconds for information about the disk drives? It looks like the table has about a half million rows and is growing by 3 rows every 5 seconds or so.
Is this intentional or is something broken?
If its broken, how can I fix it?
Is this table indexed properly?
Is this level of granularity needed? I'm willing to accept a 5 minute instead of 5 second granularity. Can I tune this somehow to not cause this level of load just to monitor for free space?