|
From: | Daniel Maher |
Subject: | Re: [Gluster-devel] successive bonnie++ tests taking longer and longer to run (system load steadily increasing) |
Date: | Mon, 22 Mar 2010 17:15:51 +0100 |
User-agent: | Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9pre) Gecko/20100217 Shredder/3.0.3pre |
On 03/22/2010 10:44 AM, Raghavendra G wrote:
As we can see, there is not much difference in Rewrite performance with/without io-cache. Hence if there are writes to the same offsets from which reads have been done, data will always be read from back-end and overhead of io-cache can be significant. In these scenarios io-cache is not recommended. However for other cases, such as mostly read-only operations, performance boost from io-cache can be significant.
Thanks for the follow-up. It certainly lines up with what should hypothetically happen (which is a good thing).
Ultimately what all of these exercises have shown is that performance tuning needs to be done on a per-instance basis, and that there is absolutely no « one size fits all » solution.
-- Daniel Maher <dma+gluster AT witbe DOT net>
[Prev in Thread] | Current Thread | [Next in Thread] |