[Bug 191348] [mps] LSI2308 with WD3000FYYZ drives disappears after hotswapping

bugzilla-noreply at freebsd.org bugzilla-noreply at freebsd.org
Wed Oct 14 13:44:26 UTC 2015


https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191348

--- Comment #34 from Karli.Sjoberg at slu.se ---
I did it! I don´t think I can provide you with much better data than this!

Today we installed a new SC837E26-RJBOD1 around 13:20 in the log and inserted
and pulled the external SAS cable a couple of times for good measure. Then at
around 13:34 we tried inserting the first disk (WD40EZRX) to test out the
internal cabling, and then it happened:

mpssas_get_sata_identify: error reading SATA PASSTHRU; iocstatus = 0x804b
mps3: Sleeping 5 seconds after SATA ID error to wait for spinup

Just kept repeating over and over again. Tried reinserting it in bay 1 didn´t
help. So we tried pulling it out from bay 1 and insert it into bay 2 but with
the same result.

So we pulled it out and inserted a Seagate ST2000DM001 into bay 1. It didn´t
show up, bay "blocked". Reinserted it into bay 2 made it appear at 13:45. Also
tried putting it in bay 3, also worked.

Pulled that one out and inserted a Samsung HD103SI first in bay 1, didn´t work,
"blocked". Then tried in bay 2 and 3 made it appear.

Then we tried another WD40EZRX in bay 1, didn´t work. Bay 2 and 3 worked
however. We went back to the first "bad" WD40EZRX and tried bay 1, 2 and 3,
still didn´t work.

There´s also trace here of me pulling out a WD RE SAS WD4001FYYG from another
SC837E26-RJBOD1 to test with in the "blocked" bay 1 at 14:09.

Here I hypothesized that this first bad WD40EZRX permanently taints every bay
it´s inserted in so we tried reinserting a drive that previously showed up, at
least in bay 2 and 3, the ST2000DM001. Inserting it in bay 1 did nothing, as
before. But now I would have thought that it actually wouldn´t show up in bay 2
or 3 either since we had tried inserting the "bad" drive in there. But the
ST2000DM001 actually did show up there, in bay 2 and 3, that is. Bay 1 is
however "blocked" for any of the SATA drives.

As a last test we took the WD RE SAS WD4001FYYG drive and inserted it into both
bay 1 and 2 which worked, as expected, at 13:13.

Full log uploaded named "kernel-critical log"

Best Regards
Karli Sjöberg

-- 
You are receiving this mail because:
You are the assignee for the bug.


More information about the freebsd-bugs mailing list