Folks,
Markus pointed out today that osgeo1 (trac/etc) was pretty much unresponsive.
I tried restarting apache but that only helped a bit temporarily. Eventually
I killed the dmsc (peer1 automated backup) process and things came back to
normal.
I did not restart dmsc, and I have no idea how to do so.
Markus suggested we run it with a somewhat higher nice level, which might
help with contention. I wonder if there is anything we can do to reduce the
frequency or completeness of backups it runs to reduce io load.
However, I don't know anything much about dmsc, and I'm not keen to learn
a lot. Is there any other primary admin interested in digging into this?
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | Geospatial Programmer for Rent