Conifer restart this morning (10:00 am)

[Daniel Scott]

Around 9:50 AM, the system started giving timeout errors due to very high database load.

I restarted all services shortly after 10:00 AM and things seem to be running smoothly again. Let me know if you discover otherwise!

Thank you, and sorry for the disruption,

Dan

--

Dan Scott
Associate Librarian & Chair, Library and Archives
Bibliothécaire agrégé & Directeur, Bibliothèque et archives
Laurentian University / Université Laurentienne
Phone: 705-675-1151 x3315

[Daniel Scott]

Whatever unusual activity triggered the high database load starts happening again shortly after I restart the services.

I'm going to shut things down again, and bring them back up. However, I won't enable the Web Staff Client, the reporter service, or the Z39.50 servers to begin with, just to see if I can rule those out.

[Daniel Scott]

Okay. So far the restart seems to be holding steady (knock wood, fingers crossed, etc). I've also shut down a few more bots that were turning up in the logs for good measure.

I'm going to start re-enabling services one at a time, starting with the reports generator at 11:15; then the Z39.50 server at 11:30; then the Web Staff Client at 11:45, to see if we can narrow down where the source of the problem might be.

[Daniel Scott]

Well, we're bogging down again. I have logs so I'm going to try to determine where the problems start showing up.

Sorry folks.

[Daniel Scott]

Okay, I think I figured it out. Some of our libraries had their hours of operation set to closed for every day of the week in anticipation of the upcoming holidays.

This caused the date due calculation to freak out, as it effectively was trying to calculate to infinity for every circulation transaction.

I've set some hours of operation for every library as a placeholder and will update them again on the other side of the holidays.

Services should be back to normal (for some definition of normal) now. I will continue to monitor it.