[Top] [All Lists]

Re: [MMTests] IO metadata on XFS

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: [MMTests] IO metadata on XFS
From: Mel Gorman <mgorman@xxxxxxx>
Date: Tue, 3 Jul 2012 12:44:24 +0100
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, linux-mm@xxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, linux-fsdevel@xxxxxxxxxxxxxxx, xfs@xxxxxxxxxxx, dri-devel@xxxxxxxxxxxxxxxxxxxxx, Keith Packard <keithp@xxxxxxxxxx>, Eugeni Dodonov <eugeni.dodonov@xxxxxxxxx>, Daniel Vetter <daniel.vetter@xxxxxxxx>, Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
In-reply-to: <20120703105951.GB14154@xxxxxxx>
References: <20120620113252.GE4011@xxxxxxx> <20120629111932.GA14154@xxxxxxx> <20120629112505.GF14154@xxxxxxx> <20120701235458.GM19223@dastard> <20120702063226.GA32151@xxxxxxxxxxxxx> <20120702143215.GS14154@xxxxxxx> <20120702193516.GX14154@xxxxxxx> <20120703001928.GV19223@dastard> <20120703105951.GB14154@xxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Tue, Jul 03, 2012 at 11:59:51AM +0100, Mel Gorman wrote:
> > Can you run latencytop to see
> > if there is excessive starvation/wait times for allocation
> > completion?
> I'm not sure what format you are looking for.  latencytop is shit for
> capturing information throughout a test and it does not easily allow you to
> record a snapshot of a test. You can record all the console output of course
> but that's a complete mess. I tried capturing /proc/latency_stats over time
> instead because that can be trivially sorted on a system-wide basis but
> as I write this I find that latency_stats was bust. It was just spitting out
> Latency Top version : v0.1
> and nothing else.  Either latency_stats is broken or my config is. Not sure
> which it is right now and won't get enough time on this today to pinpoint it.

PEBKAC. Script that monitored /proc/latency_stats was not enabling
latency top via /proc/sys/kernel/latencytop

Mel Gorman

<Prev in Thread] Current Thread [Next in Thread>