| To: | linux-kernel@xxxxxxxxxxxxxxx |
|---|---|
| Subject: | Does anyone have a benchmark of chunk sizes with SW RAID5? |
| From: | Justin Piszcz <jpiszcz@xxxxxxxxxxxxxxx> |
| Date: | Tue, 26 Jun 2007 10:44:10 -0400 (EDT) |
| Cc: | xfs@xxxxxxxxxxx |
| Sender: | xfs-bounce@xxxxxxxxxxx |
I am running one with 64,128,256,512,1024,2048,4096,8192,16384,32768,65536
it it accepts up that high.. With XFS and _ALL_ default mount and
filesystem settings to target this specific tunable.p34:~# /usr/bin/time ./benchraid.sh Tue Jun 26 10:25:21 EDT 2007: Creating RAID5 array with 64 chunk size... Tue Jun 26 10:25:21 EDT 2007: Increasing RAID rebuild speed... Tue Jun 26 10:25:21 EDT 2007: RAID is still building... Tue Jun 26 10:35:21 EDT 2007: RAID is still building... |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: Wrong Data Pointer-XFS File system, Sandy1 |
|---|---|
| Next by Date: | Re: [PATCH 4/7][TAKE5] support new modes in fallocate, Andreas Dilger |
| Previous by Thread: | 128k proved to be optimal for max_sectors_kb, Justin Piszcz |
| Next by Thread: | reasonable xfs fs size for 30-100 TB?, Ralf Gross |
| Indexes: | [Date] [Thread] [Top] [All Lists] |