... .4 actually buffers write data on > RAID-1 > which is inherently bad (after all if I do RAID-1 it's for the sake ... in 2.4 ? The basic performance should be the same, and ... s free Mandriva Linux release 2008.1 (Cooker) for i586 -- To unsubscribe ...

linux-kernel - "J.A. Magallón" - Mar 26 2008 - 04:25

... > I/O on a RAID-1 setup for 2.4 versus 2 ... dd if=/dev/zero of=/mnt/raid/testdd01 bs=1M count=1024 ... if=/dev/zero of=/mnt/raid/testdd01 bs=1M count=1024 \ ... if=/dev/zero of=/mnt/raid/testdd02 bs=1M count=1024 ... if=/dev/zero of=/mnt/raid/testdd02 bs=1M count=1024 ...

linux-kernel - Emmanuel Florac - Mar 26 2008 - 08:36

On Wed, Mar 26, 2008 at 1:36 PM, Emmanuel Florac RAID-1 setup for 2.4 versus ... usually increases performance, as well as increasing the read and write ... be helpful when optimizing NFS performance. Bart. -- To unsubscribe from this ...

linux-kernel - Bart Van Assche - Mar 26 2008 - 09:22

... >> I/O on a RAID-1 setup for 2.4 versus 2 ... if=/dev/zero of=/mnt/raid/testdd01 bs=1M count=1024 ... if=/dev/zero of=/mnt/raid/testdd01 bs=1M count=1024 ... if=/dev/zero of=/mnt/raid/testdd02 bs=1M > count=1024 ... if=/dev/zero of=/mnt/raid/testdd02 bs=1M count=1024 ...

linux-kernel - Bill Davidsen - Mar 27 2008 - 18:03

... it looks like 2.4 actually buffers write data on RAID-1 which is ... I do RAID-1 it's for the sake of data integrity, and write caching just counters that). However, ... . Though I can substantially enhance 2.6 performance through tweaking (playing with read ...

linux-kernel - Emmanuel Florac - Mar 26 2008 - 04:05

... software RAID-1 (tested >>>>> 2.4.32 >>>>> and 2.4.36 ... in the kernel. With all the mods and queued i/o ... O patterns you're instructing it to perform are not optimized ... dd absolutely killed real application performance. -- Chris -- To unsubscribe from this ...

linux-kernel - Chris Snook - Mar 26 2008 - 12:39

... like 2.4 actually buffers write data on RAID-1 > which ... RAID-1 it's for the sake of > data integrity, and ... I can substantially enhance 2.6 performance through tweaking > (playing with ... difference between your small systems and your large systems is the ...

linux-kernel - Chris Snook - Mar 26 2008 - 12:51

... Linux software RAID-1 (tested >>>> 2.4.32 >>>> and 2.4.36.2 ... 2.6.18.8, 2.6.22.16, 2.6.24.3) ... video capture, streaming video servers, and similar. >> > > dd uses unaligned stack- ... care about streaming I/O performance use large, > aligned buffers which ...

linux-kernel - Bill Davidsen - Mar 25 2008 - 19:42

... subsystems at the same time: RAID-1 + the filesystem on top of ... you want to test RAID-1 performance you should specify a ... of=... instead of a file (and the direct I/O flags). There are considerable performance differences between filesystems for large ...

linux-kernel - Bart Van Assche - Mar 26 2008 - 04:42

... software RAID-1 (tested 2.4.32 > and 2.4.36. ... machines (same hard drives though) and it remained consistent across > the ... effect much more pronounced on 2.4 (i.e. writing goes ... trend has unfortunately been slower, and responses demanding you use another ...

linux-kernel - Bill Davidsen - Mar 25 2008 - 18:37

... software RAID-1 (tested 2.4.32 >>> and 2.4.36. ... effect much more pronounced on 2.4 (i.e. writing goes ... video capture, streaming video servers, and similar. > dd uses unaligned stack ... care about streaming I/O performance use large, aligned buffers which ...

linux-kernel - Chris Snook - Mar 25 2008 - 19:13

... > I'm using XFS usually, and I've also checked against the raw devices > and it looks the same (2.4 still faster). ... to different buffering policy in the RAID subsystem. You are welcome to post ... with dd for direct I/O on a RAID-1 setup for 2.4 versus ...

linux-kernel - Bart Van Assche - Mar 26 2008 - 07:15

... software RAID-1 (tested >>>> 2.4.32 >>>> and 2.4.36. ... video capture, streaming video servers, and similar. >> > > dd uses unaligned stack- ... care about streaming I/O performance use large, > aligned buffers which ... advice above (on ext3 on 2.6.24.5-85.fc8 ...

linux-kernel - Pádraig Brady - Jul 16 2008 - 10:52

... software RAID-1 (tested 2.4.32 > and 2.4.36. ... machines (same hard drives though) and it remained consistent across > the board, with /mountpoint a software RAID-1. > Actually checking disk activity ... effect much more pronounced on 2.4 (i.e. writing goes ...

linux-kernel - Chris Snook - Mar 25 2008 - 18:00

... software RAID-1 (tested 2.4.32 > > and 2.4.36. ... is way > > faster than 2.6 ( tested 2.6.17.13, ... machines (same hard drives though) and it remained consistent across > > the ... board, with /mountpoint a software RAID-1. > > Actually checking disk activity ...

linux-kernel - Bart Van Assche - Mar 26 2008 - 03:15

... two subsystems at the same > time: RAID-1 + the filesystem on top of it. I'm using XFS usually, and I've also checked against the raw devices and it looks the same (2.4 still faster). ... using a single disk drive vs. RAID-1, obviously due to different ...

linux-kernel - Emmanuel Florac - Mar 26 2008 - 07:07

... software RAID-1 (tested >>>>> 2.4.32 >>>>> and 2.4.36 ... care about streaming I/O performance use large, >> aligned buffers which ... with some architectures than others 2) block alignment, which is necessary ... for direct I/O You're on the right track with ...

linux-kernel - Chris Snook - Jul 16 2008 - 14:18

... software RAID-1 (tested 2.4.32 >> and 2.4.36. ... is way >> faster than 2.6 ( tested 2.6.17.13 ... machines (same hard drives though) and it remained consistent across >> the ... video capture, streaming video servers, and similar. -- Bill Davidsen

linux-kernel - Bill Davidsen - Mar 25 2008 - 18:47

... to be portable, that are not rewritten and fancy > optimized to take advantage of latest ... that in 2.4 ? The basic performance should be the same, > and all ... making useful) more typical desktop configurations. And yes, these are "real" applications, and ...

linux-kernel - Bill Davidsen - Mar 27 2008 - 17:49

... Or, even better, you can use xdd with > the flags -dio -processlock. > No, you want your benchmark to measure performance doing what the application does. Do unless you have an application which has been heavily Linux-ized you don't want to measure ...

linux-kernel - Bill Davidsen - Mar 27 2008 - 17:53