On Tue, 10 Mar 2009 16:48:09 -0700 Ben Greear <greearb@xxxxxxxxxxxxxxx> wrote: > FUJITA Tomonori wrote: > > On Tue, 10 Mar 2009 15:34:08 -0700 > > Ben Greear <greearb@xxxxxxxxxxxxxxx> wrote: > > > >> I wrote a script to create lots of iscsi volumes on loop devices. > >> > >> Seems to run fine up to 111, and then tgtd crashes and tgtadm > >> gets a buffer overflow. > >> > >> The script I used to create the problem and a capture of the > >> crash is attached. > >> > >> I'm using standard RPMs on Fedora 10, kernel > >> Linux iSCSi-test 2.6.27.15-170.2.24.fc10.x86_64 #1 SMP Wed Feb 11 23:14:31 EST 2009 x86_64 x86_64 x86_64 GNU/Linux > >> > >> Let me know if I can provide any additional info. > > > > Your subject gave me an impression that you are creating lots of > > logical units with one target but looks like you are trying to create > > lots of targets with a single logical unit, right? > > I'm new to iscsi, so not so sure of my terminology, but I think > you have it correct. I'm working on a test tool that uses open-iscsi > to mount lots of volumes on a single machine, to be used to load-test > iscsi target systems. I was using tgtd as the iscsi target for > testing my initiator setup. > > > We use pthread and glibc has the limit of the number of > > pthreads. Looks like you hit it. I guess that tgt also has a bug (tgt > > should handle pthread creation failure or something properly) though. > > Looks like you can have many more, if this is accurate, and that's > from 2004: > http://forum.soft32.com/linux2/Thread-count-limit-ftopict14647.html I think that 2.4.X kernel's way to handle pthread is different from 2.6.X. So I'm not sure this can be applied to the current situation. FYI, we consume 2 process + 8 pthread per one target with a single logica unit. -- To unsubscribe from this list: send the line "unsubscribe stgt" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
