Re: [zfs-discuss] Filebench Performance is weird

2010-03-03 Thread Richard Elling
On Mar 2, 2010, at 9:43 PM, Abdullah Al-Dahlawi wrote: > Greeting Richard > > After spending alomost 48 hours working on this problem, I believe I've > discovered the BUG in Filebench !!!. > > I do not believe it is the change directory that you have indicated below > cause this directory is u

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Abdullah Al-Dahlawi
Greeting Richard After spending alomost 48 hours working on this problem, I believe I've discovered the BUG in Filebench !!!. I do not believe it is the change directory that you have indicated below cause this directory is used to dump the stat data at the end of the benchmarks, it is NOT used d

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Abdullah Al-Dahlawi
Hi Richard Are you saying that running filebench with a configuration file is different (it flushes the cache) from running it as an interactive mode for the same exact workload configuration If the answer is yes , how can I make filebench behave the same while running in NON-interactive mod

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Richard Elling
see below... On Mar 2, 2010, at 12:38 AM, Abdullah Al-Dahlawi wrote: > Greeting All > > I am using Filebench benchmark in an "Interactive mode" to test ZFS > performance with randomread wordload. > My Filebench setting & run results are as follwos > -

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Abdullah Al-Dahlawi
Hi Phil Good notice , but believe me this is not the problem . I just cut & pasted the filebench "config" file after I rerun it with in shorter time, the orginal one was SAME with 100 seconds. and by the way I always enmpty the cache with zfs export & import before every run Filebench still beh

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Phil Harman
I see at least two differences: 1. duration 30s vs 100s (so not "SAME") 2. your manual test doesn't empty the cache Of course, it is the latter that makes all the difference. Hope this helps, Phil Sent from my iPhone On 2 Mar 2010, at 08:38, Abdullah Al-Dahlawi wrote: Greeting All I am u

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Abdullah Al-Dahlawi
Hi My pool is very simple disk partition on a workstation abdul...@hp_hdx_16:~# zpool list NAMESIZE USED AVAILCAP HEALTH ALTROOT hdd19.6G 5.00G 14.6G25% ONLINE - rpool 123G 7.82G 115G 6% ONLINE - just testing randomread in filebench . On Tue, Mar 2, 201

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Robert Milkowski
On 02/03/2010 08:38, Abdullah Al-Dahlawi wrote: Greeting All I am using Filebench benchmark in an "Interactive mode" to test ZFS performance with randomread wordload. My Filebench setting & run results are as follwos -

Re: [zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Will Murnane
On Tue, Mar 2, 2010 at 08:38, Abdullah Al-Dahlawi wrote: > I am using Filebench benchmark in an "Interactive mode" to test ZFS > performance with randomread wordload. What's your pool configuration? Are you just load testing, or do you have an application in mind for this system? Will __

[zfs-discuss] Filebench Performance is weird

2010-03-02 Thread Abdullah Al-Dahlawi
Greeting All I am using Filebench benchmark in an "Interactive mode" to test ZFS performance with randomread wordload. My Filebench setting & run results are as follwos -- filebench> set $filesize=5g filebench>