[dm-devel] v4.8 dm-mpath

[Bart Van Assche]

On 08/16/2016 07:43 PM, Mike Snitzer wrote:
> On Tue, Aug 16 2016 at  3:12pm -0400,
> Mike Snitzer <snitzer at redhat.com> wrote:
>
>> On Tue, Aug 16 2016 at  1:32pm -0400,
>> Bart Van Assche <bart.vanassche at sandisk.com> wrote:
>>
>>> Can you have a look at this?
>>
>> I'll get linux-dm.git's 'dm-4.8' branch (v4.8-rc2) loaded on one of my
>> testbed systems to run the mptest testsuite.  It should provide coverage
>> for simple failover and failback.
>
> I successfully ran mptest against that v4.8-rc2 based kernel.  Any
> chance you could try mptest to see if it also passes for you?  Or if it
> is able to trigger the issue for you.
>
> mptest's runtest script defaults to setting up scsi-mq and dm-mq.

Hello Mike,

mptest on top of scsi_debug passes on v4.8-rc3 on the same system I used 
to run my SRP tests:

# ./runtest
[ ... ]
PASSED:  test_00_no_failure test_01_sdev_offline test_02_sdev_delete 
test_03_dm_failpath test_04_dm_switchpg
FAILED:

My SRP tests pass on the same setup if I disable scsi-mq. However, I 
still sporadically see I/O errors if I run srp-test 02 on kernel 
v4.8-rc3 with scsi-mq and dm-mq:
# srp-test/run_tests -t 02
[ ... ]
verify: bad magic header 0, wanted acca at file 
/root/mnt1/data-integrity-test.1.0 offset 581632, length 4096
fio: io_u error on file /root/mnt1/data-integrity-test.8.0: Input/output 
error: write offset=8888320, buflen=4096

Bart.