Hi,
That's probably true. zil_itx_cleanup may have to move a lot of entries to
another list to free up. It looks your test is related to below bugs. But I
think they should have been fixed. Can you check your version of
solaris/opensolaris to see whether these patch has been included.
Bob,
Thanks for your reply. As you mentioned, adjusting the zfs tunables to
reduce the transaction group size yields shorter but more frequent pauses.
Unfortunately, this workaround doesn't sufficiently meet our needs (pauses
are still too long).
I've reviewed the forum archives and read a
On Fri, 26 Feb 2010, Shane Cox wrote:
I've reviewed the forum archives and read a number of threads related to this
issue. However I
didn't find a root-cause explanation for these pauses, only talk of how to
ameliorate them. In my
particular case, I would like to know why zfs_log_writes
Hi,
This page may indicate the root cause.
http://blogs.sun.com/roch/entry/the_new_zfs_write_throttle
ZFS will throttle the write speed to match the write speed to the txg to the
speed of DISK IO. If it detects the modest measure(1 tick pause) cannot
prevent the tx group from being too large, it
Thanks for your reply. I disabled write throttling, but didn't observe any
change in behavior. After doing some more research, I have a theory as to
the root cause of the pauses that I'm observing.
Near the end of spa_sync, writes are blocked in function zil_itx_assign as
illustrated by the
On Thu, 25 Feb 2010, Shane Cox wrote:
I'm new to ZFS and looking for some assistance with a performance problem:
At the interval of zfs_txg_timeout (I'm using the default of 30), I observe
100-200ms
pauses in my application. Based on my application log files, it appears that
the
write()
