<div class="gmail_quote">On Tue, May 5, 2009 at 6:15 PM, Lars Ellenberg <span dir="ltr"><<a href="mailto:lars.ellenberg@linbit.com">lars.ellenberg@linbit.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="im">On Tue, May 05, 2009 at 03:57:16PM -0400, Gennadiy Nerubayev wrote:<br>
> On Sat, May 2, 2009 at 2:06 PM, Lars Ellenberg <<a href="mailto:lars.ellenberg@linbit.com">lars.ellenberg@linbit.com</a>>wrote:<br>
><br>
> > nice. you are going into page cache here, mostly<br>
> > (how much RAM did you say you have?)<br>
> ><br>
><br>
> 4GB on one node, 8GB on the other. I verified that benchmark numbers are the<br>
> same on both nodes.<br>
><br>
> > dd if=/dev/zero of=/dev/drbd0 bs=4M count=1000 oflag=direct<br>
> > > 4194304000 bytes (4.2 GB) copied, 17.6633 seconds, 237 MB/s<br>
> ><br>
> > do variations.<br>
> > use 512k, 1M, 2M, 100M.<br>
> ><br>
><br>
> After rerunning these various sizes, there is no appreciable performance<br>
> difference from bs=4M, in either connected or disconnected mode, across all<br>
> flags. I didn't do connected benchmarks on a single core, as the resync<br>
> performance dropped by 100MB/s with only one core..<br>
<br>
</div>Ah. right.<br>
in that case, try "syncer { cpu-mask 3; }" and see if that improves<br>
performance during normal operation on your dual core cpu.</blockquote><div><br>Not much effect, except dsync which appears to be a little faster by about 10MB/s.<br><br>-Gennadiy</div></div>