Messages in this thread | | | Date | Wed, 23 Jun 2010 12:02:52 +0200 | From | Christoph Hellwig <> | Subject | Re: trying to understand READ_META, READ_SYNC, WRITE_SYNC & co |
| |
On Mon, Jun 21, 2010 at 04:25:04PM -0400, Vivek Goyal wrote: > In my testing in the past, this was helping if lots of sequential readers > are running in a system (with 100ms slice each) and if there is another > reader doing small file reads. Without meta data based preemption check, > latency of opening small file used to be very high (Because all the sync > sequntial reader gets to consume their 100ms slices first). > > Situation should be somewhat better now after corrado's changes of > reducing slice length dynamically. But I suspect that we will still > experience significantly reduced latency for small file reads in presece > of multiple sequntial reads going on.
Any chance we could create a benchmark suite for the I/O schedulers? I'd really like to get these annotations correct, and having an automated suite that shows the numbers for the interesting workloads would help greatly with that.
| |