2011-06-11 20:34, Jim Klimov пишет:
time dtrace -n 'syscall::*write:entry /fds[arg0].fi_fs == "zfs"/ {
@[execname,fds[arg0].fi_pathname]=count(); }'
This time I gave it more time, and used the system a bit -
this dtrace works indeed, but there are still too few file
accesses:
# time dtrace -n 'syscall::*write:entry /fds[arg0].fi_fs == "zfs"/ {
@[execname,fds[arg0].fi_pathname]=count(); }'
dtrace: description 'syscall::*write:entry ' matched 2 probes
^C
svc.startd
/var/svc/log/network-iscsi-initiator-dcpool:default.log 3
svc.startd
/var/svc/log/network-iscsi-target-dcpool:default.log 3
bash
/root/.bash_history 4
sshd
/var/adm/utmpx 4
syslogd
/var/adm/messages 6
sshd
/var/adm/lastlog 8
sshd
/var/adm/wtmpx 8
fmd
/var/fm/fmd/infolog_hival 14
svc.configd
/etc/svc/repository.db 17
svc.configd
/etc/svc/repository.db-journal 22
freeram-watchdog
/var/log/freeram-watchdog.log.1307810438 177
real 2m59.175s
user 0m1.474s
sys 0m0.349s
However I've also rebooted the system and did not import
the iSCSI device nor the "dcpool" I'm trying to repair -
to rule this out - these 2Mb/s writes were still there.
Then I played with TXG sync times, and it seems evident
now that one TXG sync on my rpool takes 2-3Mb of written
data with whatever frequency I've set. Kind of quite a
lot to update a couple of blocks or so!..
//Jim
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
