Performance statistics - 1.3.7 slowness

[John A. Murdie]

Is there anything in CUPS that will give me scheduler performance statistics in real-time? (I'm not referring to user page counts - that's there already, e.g. with PyKota or something similar; we use a home-grown one to meet our  own specific needs.) I could, of course, dash something off that would analyse access_log and error_log.

We have two CUPS schedulers here, on the same host - with various tweaks to keep them apart. As I write, one is speedy and behaving normally, but the other is glacial. The host is hardly loaded at all. We've been running 1.3.7/Kerberos in 'production' for two weeks - simple-minded testing before that revealed no problems we couldn't fix. It's been mostly fine up to now, but two afternoons ago this scheduler became terribly slow. I restarted it, and all then seemed to be well. This afternoon, the slowness has returned, and restarting it has made no improvement. It's difficult to see how many print jobs are arriving a minute - not a large number compared with what I gather is normal for other installations out there. I'd say we have perhaps 50-100 users of each print scheduler.

A statistics display would ideally give me an idea of what was causing the problem. Any other suggestions?

John A. Murdie