[stgt] tgtd memory usage

FUJITA Tomonori fujita.tomonori at lab.ntt.co.jp
Sun Mar 27 11:08:10 CEST 2011


On Fri, 25 Mar 2011 09:54:05 +0100
Tomasz Chmielewski <mangoo at wpkg.org> wrote:

> On 24.03.2011 02:45, FUJITA Tomonori wrote:
> > On Wed, 23 Mar 2011 17:37:53 +0300
> > Степан Фёдоров<stepan.fedorov at clodo.ru>  wrote:
> >
> >> My iSER target node has 24Gb of memory. When i add about ~500 LUNs
> >> with tgtadm, next i got error "tgtadm: out of memory".
> >>
> >> 24Gb/500 LUNs == "something about 50 megabytes of memory per LUN".
> >>
> >> My question is: how can i control this memory usage, and is there
> >> ability for me to add about 3500 LUNs on my node with 24Gb of memory?
> >
> > What version of tgt you use? x86-64 or x86?
> >
> > I suspect that you hit the number of threads that a single process can
> > have.
> >
> > The amount of system memory can't do anything for it.
> 
> What are the options to have it worked around (long term, in tgtd code)?
> 
> Is preforking a new tgtd process if X threads are reached an option?

I've not tried but sounds like too much memory than I expected. I need
to investigate.

About the number of threads, we creates several threads per lun (they
perform I/Os). Probably, we should has the fixed number of pthreads to
perform I/Os.
--
To unsubscribe from this list: send the line "unsubscribe stgt" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html



More information about the stgt mailing list