On Thu, 2016-02-11 at 18:49 -0500, Dan Lane wrote: > On Thu, Feb 11, 2016 at 2:48 AM, Nicholas A. Bellinger > <nab@xxxxxxxxxxxxxxx> wrote: > > Hello Dan, > > > > On Wed, 2016-02-10 at 21:30 -0500, Dan Lane wrote: > >> > > >> > SUCCESS! > >> > > >> > The latest changes have the filer working stable, I just benchmarked a Win7 > >> > VM on an ESXi host and hit 400+MB/s without any crashing! > >> > > > > > Thanks for the update. > > > >> > I still see a fair number of the following errors in messages, I'm not sure > >> > if it's something to worry about or not, especially considering these > >> > numbers. > >> > > >> > Feb 10 21:18:28 dracofiler kernel: ABORT_TASK: Found referenced qla2xxx > >> > task_tag: 1176876 > >> > Feb 10 21:18:28 dracofiler kernel: ABORT_TASK: Sending > >> > TMR_TASK_DOES_NOT_EXIST for ref_tag: 1176876 > >> > Feb 10 21:18:48 dracofiler kernel: ABORT_TASK: Found referenced qla2xxx > >> > task_tag: 1199888 > >> > Feb 10 21:18:48 dracofiler kernel: ABORT_TASK: Sending > >> > TMR_TASK_DOES_NOT_EXIST for ref_tag: 1199888 > >> > Feb 10 21:19:07 dracofiler kernel: ABORT_TASK: Found referenced qla2xxx > >> > task_tag: 1212868 > >> > Feb 10 21:19:07 dracofiler kernel: ABORT_TASK: Sending > >> > TMR_TASK_DOES_NOT_EXIST for ref_tag: 1212868 > >> > Feb 10 21:20:19 dracofiler kernel: ABORT_TASK: Found referenced qla2xxx > >> > task_tag: 1188008 > >> > Feb 10 21:20:19 dracofiler kernel: ABORT_TASK: Sending > >> > TMR_TASK_DOES_NOT_EXIST for ref_tag: 1188008 > >> > Feb 10 21:20:19 dracofiler kernel: ABORT_TASK: Found referenced qla2xxx > >> > task_tag: 1188052 > >> > Feb 10 21:20:19 dracofiler kernel: ABORT_TASK: Sending > >> > TMR_TASK_DOES_NOT_EXIST for ref_tag: 1188052 > >> > > >> > Thanks again for all the help you provided. I helped a friend with a > >> > similar setup get his fixed as well and he had the same results. > > > > Keep in mind your target-pending/4.4-stable branch is still missing the > > active I/O remote port LUN_RESET + session disconnect bug-fix currently > > being tested in target-pending/master here: > > > > https://git.kernel.org/cgit/linux/kernel/git/nab/target-pending.git/commit/?id=0f4a943168f31d29a1701908931acaba518b131a > > > >> Let me know if you need any further testing of the qlogic stuff. > >> > > > > > Let's have look your fc_host class side stats to verify FC physical > > layer for tcm_qla2xxx ports are working as expected. > > > > head /sys/class/fc_host/host*/statistics/* > > > > Wow, lots of info... Here you go! > BTW, I have two QLE2462 cards in my storage box, one has two > connections to a single host and the other connects to two FC switches > in a blade chassis. So AFAICT, nothing looks out of the ordinary wrt to the stats counters. To further debug, I'd recommend looking at the stats counters on your FC switch and on the ESX FC host generating constant ABORT_TASKS to determine who is responsible for dropping packets. -- To unsubscribe from this list: send the line "unsubscribe target-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
