Oops. Typed that command wrong. Sorry.
Should be
# ps -o vsz,args -p `pgrep idmapd`
# svcadm refresh idmap
# ps -o vsz,args -p `pgrep idmapd`
Missed the -p.
Jordan Brown wrote:
I've been investigating this leak and believe I have a fix in hand.
If it's the leak I'm looking at, it will leak proportional to the number
of domains, every time that idmapd does a discovery run. For my system
that's every five minutes.
Try this sequence. If it's the leak I'm looking at, it'll leak hundreds
of kilobytes on each svcadm refresh.
# ps -o vsz,args `pgrep idmapd`
# svcadm refresh idmap
# ps -o vsz,args `pgrep idmapd`
roland wrote:
i can actually "watch" idmapd leak memory - i.e. the leaking happens
in specific time-interval .
when i attach via truss to the process, i see idmapd leaking whenever
it wakes up from sleep, i.e. nothing happens for most of the time, but
then the process wakes up and does something, truss spits out a lot of
stuff and immediately is see vsz go up.
_______________________________________________
cifs-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/cifs-discuss
_______________________________________________
cifs-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/cifs-discuss
