| To: | xfs@xxxxxxxxxxx |
|---|---|
| Subject: | Fw: Re: XFS mkfs/mount options |
| From: | Mark <musicman529@xxxxxxxxx> |
| Date: | Wed, 18 Jun 2008 00:21:01 -0700 (PDT) |
| Domainkey-signature: | a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Received:X-Mailer:Date:From:Reply-To:Subject:To:MIME-Version:Content-Type:Message-ID; b=olE9+g4I4R5+FNFwppU1s6gwLh10iRlvV47wgRUj+CcN/N5UC7N8hGbIunT28wZZCLC1Z9QmXYP+tACh1EMAiyCzXYgyYw42dMtjhAhPR/uVaeuL+dvjD8oIEg7GdD+xWsMsR6bxxm0TvQoKmh6hGaYC5Gbb7toNIGhMQ1Wt5A8=; |
| Reply-to: | MusicMan529@xxxxxxxxx |
| Sender: | xfs-bounce@xxxxxxxxxxx |
To all,
It appears I have found a defect in my testing. Generally, the procedure has
been something like:
create filesystem
mount filesystem and set owner/group
run benchmark
alter some condition (CPU speed or IRQ routing)
re-run benchmark
compare results
The second run is contaminated by caching effects from the first run. This
accounts for the roughly 5% speedup I saw after enabling explicit IRQ routing.
Once I corrected this oversight, the difference became statistical noise.
My apologies to all for this red herring.
--
Mark
"What better place to find oneself than
on the streets of one's home village?"
--Capt. Jean-Luc Picard, "Family"
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH] fix XFSQA 144, Donald Douwsma |
|---|---|
| Next by Date: | Fw: 2.6.26-rc6 link count mismatch for inode, Marco Berizzi |
| Previous by Thread: | Re: XFS mkfs/mount options, Martin Steigerwald |
| Next by Thread: | E-MEDISYS'08: Extended Paper submission deadline, E-MEDISYS 08 |
| Indexes: | [Date] [Thread] [Top] [All Lists] |