Re: [zfs-discuss] iSCSI access patterns and possible improvements?

2013-01-18 Thread Richard Elling
On Jan 18, 2013, at 4:40 AM, Jim Klimov wrote: > On 2013-01-18 06:35, Thomas Nau wrote: If almost all of the I/Os are 4K, maybe your ZVOLs should use a volblocksize of 4K? This seems like the most obvious improvement. >>> >>> 4k might be a little small. 8k will have less metadata ove

Re: [zfs-discuss] iSCSI access patterns and possible improvements?

2013-01-18 Thread Richard Elling
On Jan 17, 2013, at 9:35 PM, Thomas Nau wrote: > Thanks for all the answers more inline) > > On 01/18/2013 02:42 AM, Richard Elling wrote: >> On Jan 17, 2013, at 7:04 AM, Bob Friesenhahn > > wrote: >> >>> On Wed, 16 Jan 2013, Thomas Nau wrote: >>> Dea

Re: [zfs-discuss] Heavy write IO for no apparent reason

2013-01-18 Thread Timothy Coalson
On Fri, Jan 18, 2013 at 4:55 PM, Freddie Cash wrote: > On Thu, Jan 17, 2013 at 4:48 PM, Peter Blajev wrote: > >> Right on Tim. Thanks. I didn't know that. I'm sure it's documented >> somewhere and I should have read it so double thanks for explaining it. >> > > When in doubt, always check the ma

Re: [zfs-discuss] Heavy write IO for no apparent reason

2013-01-18 Thread Freddie Cash
On Thu, Jan 17, 2013 at 4:48 PM, Peter Blajev wrote: > Right on Tim. Thanks. I didn't know that. I'm sure it's documented > somewhere and I should have read it so double thanks for explaining it. > When in doubt, always check the man page first: man zpool It's listed in the section on the "

Re: [zfs-discuss] Heavy write IO for no apparent reason

2013-01-18 Thread Peter Blajev
Right on Tim. Thanks. I didn't know that. I'm sure it's documented somewhere and I should have read it so double thanks for explaining it. -- Peter Blajev IT Manager, TAAZ Inc. Office: 858-597-0512 x125 On Thu, Jan 17, 2013 at 4:18 PM, Timothy Coalson wrote: > On Thu, Jan 17, 2013 at 5:33 PM,

Re: [zfs-discuss] iSCSI access patterns and possible improvements?

2013-01-18 Thread Jim Klimov
On 2013-01-18 06:35, Thomas Nau wrote: If almost all of the I/Os are 4K, maybe your ZVOLs should use a volblocksize of 4K? This seems like the most obvious improvement. 4k might be a little small. 8k will have less metadata overhead. In some cases we've seen good performance on these workload