Re: [perf-discuss] zfs write slow

2010-02-01 Thread Anatoly
Thank you very much! I will try to lower IO queue, disable arc cache, and disable cache flushing and post back results. Logbias property sounds interesting. It's sad that it's not implemended in solaris yet. -- This message posted from opensolaris.org

Re: [perf-discuss] zfs write slow

2010-02-01 Thread Richard Elling
On Jan 31, 2010, at 9:39 PM, Anatoly wrote: > Yes, it has battery-backed cache and it's working fine. > I'm worried about fact that simple > dd if=/dev/zero of=/zpool1 bs=8192k count=10240 > gives 40MB/s in zpool iostat > > Maybe Oracle settings are wrong? > Current settings are: > filesystemio_

Re: [perf-discuss] zfs write slow

2010-02-01 Thread Paul van den Bogaard
Hi ZFS has no equivalent for directio mounting. Did you try the set zfs:zfs_nocacheflush = 1 option? (see http://www.solarisinternals.com/wiki/index.php/ZFS_Evil_Tuning_Guide#Tuning_ZFS_for_Database_Performance) Cheers Paul. On Feb 1, 2010, at 6:39 AM, Anatoly wrote: > > > gives 40MB/

Re: [perf-discuss] zfs write slow

2010-01-31 Thread Anatoly
Yes, it has battery-backed cache and it's working fine. I'm worried about fact that simple dd if=/dev/zero of=/zpool1 bs=8192k count=10240 gives 40MB/s in zpool iostat Maybe Oracle settings are wrong? Current settings are: filesystemio_options=asynch db_writer_processes=2 db_file_multiblock_read_

Re: [perf-discuss] zfs write slow

2010-01-30 Thread Bob Friesenhahn
On Sat, 30 Jan 2010, Anatoly wrote: Hello! We have Solaris 10 box with latest patches connected to Storedge 6130 by FC. Storedge disks configured as RAID5 Our Oracle 10g (10.2.0.4) DB performance is terrible. I don't know anything about the Storedge 6130, but if it includes a battery-backed w

[perf-discuss] zfs write slow

2010-01-30 Thread Anatoly
Hello! We have Solaris 10 box with latest patches connected to Storedge 6130 by FC. Storedge disks configured as RAID5 Our Oracle 10g (10.2.0.4) DB performance is terrible. Write speed reported by "zpool iostat" about 5 MB/s and "iostat" showing 100% load for data pool. Redo logs and data files