On 07/29/2010 12:52 PM, FUJITA Tomonori wrote: > On Thu, 29 Jul 2010 11:59:38 +0300 > Boaz Harrosh <bharrosh@xxxxxxxxxxx> wrote: > >>> Thanks, the pthread-per-target changes were reverted, so it should be >>> stable now. tgt had been stable before the changes. Otherwise, vendor >>> guys would hit me hard, I guess. :) >> >> Hi Tomo >> >> Do you have plans to return this in the future, once debugged and stabilised? > > I want but probably it would take long. I need to work on the > clustering feature now because my company needs it. > > >> For me too, as osd-target. It gave performance boost with multiple targets on >> same machine. This is because my target is badly written and does sync handling, >> no queue, of every thing. Funny that I did not experience problems with it >> during testing, but I'm not running a production system so I might got lucky. >> >> Please tell me your plans so I know if I'll have to fix my backend, one day, or >> can get by with this fix. > > I don't know when I can rework on this. I would recommend you to fix > your issues. Thanks. The problem that it is not a real issue, only slowness with multiple targets. And the recommendation is for a single target per machine in a 1G network anyway. So it is on the todo but will take long time here as well. I was happy that for a while it was better. Will see what happens. Thanks, have a good day Boaz -- 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