performance loss with inn 2.5.x (tradspool) ?
Matija Nalis
mnalis-ml at voyager.hr
Fri Feb 3 13:39:59 UTC 2012
Hi,
I've recently upgraded (xref master) news machine from Debian
Lenny with INN 2.4.6-snapshot-20090119 to Debian Squeeze with
INN 2.5.2-2~squeeze1
I've done it before with xrefslave INN server (tradspool with
tradindex) on similar hardware, and it didn't have any performance
issues.
However, now I've upgraded xref master server (tradspool, no readers
and no overview indexes, with about 45 realtime peers) and I'm seeing
severe performance degradation (loadavg went from 2 to 6!), which
started to create problems (server does not responded to nagios
connection probes to port 119, etc):
http://bayimg.com/NAmaaaAdg
http://bayimg.com/MAMAFaADg
I've looked in config files, and it does not appear as something
important has changed much. Load seems to be mostly I/O based.
Nothing looks suspicious in log files.
As machine runs nothing else really but inn, it leaves two general
suspects: kernel (although I guess we'd by now already see some
horrible VM regressions on dozen other I/O loaded machines on squeeze
- unless it is some very specific) and inn itself (or maybe some
library inn uses, like move from libdb-4.6 to 4.8).
I've tried moving the server disks to different machine in effort to
eliminate possible hardware malfunction which happened to coincide
with OS upgrade; but the performance issue with upgraded OS remains
more or less the same (moved from dual xeon with 1GB RAM, to
singlecore opteron with 4GB RAM and better raid controller)
I'll try downgrading kernel (when production allows) to try to
eliminate kernel as suspect...
So the question for the list - have you seen any such degradations
with upgrades of INN from 2.4 to 2.5? Or do you have any idea why it
might be happening? Did something big change in INN tradspool
handling between 2.4 and 2.5?
--
Opinions above are GNU-copylefted.
More information about the inn-workers
mailing list