Re: A lot of [ext4-dio-unwrit] threads around

[Ted Ts'o]

On Mon, Feb 07, 2011 at 07:09:26PM +0200, Ozan ÃaÄlayan wrote:
> 
> This is 2.6.35.11 on a 24-core HP Proliant server with 8 mounted
> ext4 filesystems. There are 7 LXC containers which does heavy
> compiling workload during the day.
> 
> Uptime is 3 days and currently there are 216 ext4-dio-unwrit threads
> in ps aux output? Is it normal or a symptom of a problem?

It's normal; the workqueues are created statically at mount time, one
per cpu per file system.  There are some plans in train for a more
efficient way of dealing with workqueues in general on systems with
large numbers of CPU's, but for 2.6.35, yes, that's normal.

      	      	 	    		- Ted
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/