info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
mrsas(4)
6 views
Skip to first unread message
Eugene M. Zheganin
Jul 29, 2016, 9:41:27 AM7/29/16
to
Hi.
I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
controller: sometimes, all of a sudden, it reports that disk was
reattached. I'm using zfs, and a redundant pool, so, besides the fact
that it's bad by itself, everything should continue to work, but instead
my FreeBSD starts to livelock, and all disk/zfs commands stop working -
they just hang (however, I'm able to log in, launch gstat and other
utilities, and most of the services seem to be unaffected, however,
iSCSI target starts to behave weirdly too). I wrote a letter to the LSI
support, but I guess they will first ask me to try the latest driver,
which for some reason isn't commited in the tree yet. So, two questions:
- why the latest driver (6.710.11.00) isn't in the main tree, it's from
december 2015, so at least it seems like it should have been made it to
the 11.x. In 11.x I still see the 06.709.07.00-fbsd ? may be I don't
unerstand something.
- has anyone seen anything similar on mrsas ? I'm kinda open to the
ideas. To be honest, I suspect the controller and the firmware, because
with factory-default firmware version I had seen all sorts of madness,
including disk detaching after requesting SMART data with smartmontools.
This was reproducible and solved with the firmware update, but
considering the amount of fixes that each new version does contain, I
suppose there's many of the bugs that left unfixed.
Thanks.
Eugene.
_______________________________________________
freebsd...@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to "freebsd-stabl...@freebsd.org"
I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
controller: sometimes, all of a sudden, it reports that disk was
reattached. I'm using zfs, and a redundant pool, so, besides the fact
that it's bad by itself, everything should continue to work, but instead
my FreeBSD starts to livelock, and all disk/zfs commands stop working -
they just hang (however, I'm able to log in, launch gstat and other
utilities, and most of the services seem to be unaffected, however,
iSCSI target starts to behave weirdly too). I wrote a letter to the LSI
support, but I guess they will first ask me to try the latest driver,
which for some reason isn't commited in the tree yet. So, two questions:
- why the latest driver (6.710.11.00) isn't in the main tree, it's from
december 2015, so at least it seems like it should have been made it to
the 11.x. In 11.x I still see the 06.709.07.00-fbsd ? may be I don't
unerstand something.
- has anyone seen anything similar on mrsas ? I'm kinda open to the
ideas. To be honest, I suspect the controller and the firmware, because
with factory-default firmware version I had seen all sorts of madness,
including disk detaching after requesting SMART data with smartmontools.
This was reproducible and solved with the firmware update, but
considering the amount of fixes that each new version does contain, I
suppose there's many of the bugs that left unfixed.
Thanks.
Eugene.
_______________________________________________
freebsd...@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to "freebsd-stabl...@freebsd.org"
Ultima
Jul 29, 2016, 9:57:15 AM7/29/16
to
Hello, I have experienced this twice within the last month. It was not with
this same controller, however it was not the controllers fault. Random
disks will drop and reattach. For me, this was power supply problem and
both times replacement fixed the issue.
If you haven't checked into the power supply. I would suggest testing it.
Especially If it was previously working fine.
On Fri, Jul 29, 2016 at 9:41 AM, Eugene M. Zheganin <e...@norma.perm.ru>
wrote:
this same controller, however it was not the controllers fault. Random
disks will drop and reattach. For me, this was power supply problem and
both times replacement fixed the issue.
If you haven't checked into the power supply. I would suggest testing it.
Especially If it was previously working fine.
On Fri, Jul 29, 2016 at 9:41 AM, Eugene M. Zheganin <e...@norma.perm.ru>
wrote:
Miroslav Lachman
Jul 29, 2016, 10:25:41 AM7/29/16
to
Eugene M. Zheganin wrote on 07/29/2016 15:41:
> Hi.
>
> I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
> controller: sometimes, all of a sudden, it reports that disk was
> reattached. I'm using zfs, and a redundant pool, so, besides the fact
> that it's bad by itself, everything should continue to work, but instead
> my FreeBSD starts to livelock, and all disk/zfs commands stop working -
> they just hang (however, I'm able to log in, launch gstat and other
> utilities, and most of the services seem to be unaffected, however,
> iSCSI target starts to behave weirdly too). I wrote a letter to the LSI
> support, but I guess they will first ask me to try the latest driver,
> which for some reason isn't commited in the tree yet. So, two questions:
>
> - why the latest driver (6.710.11.00) isn't in the main tree, it's from
> december 2015, so at least it seems like it should have been made it to
> the 11.x. In 11.x I still see the 06.709.07.00-fbsd ? may be I don't
> unerstand something.
>
> - has anyone seen anything similar on mrsas ? I'm kinda open to the
> ideas. To be honest, I suspect the controller and the firmware, because
> with factory-default firmware version I had seen all sorts of madness,
> including disk detaching after requesting SMART data with smartmontools.
> This was reproducible and solved with the firmware update, but
> considering the amount of fixes that each new version does contain, I
> suppose there's many of the bugs that left unfixed.
mrsas is very strange so I moved to mfi driver for Dell PowerEdge T130
> Hi.
>
> I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
> controller: sometimes, all of a sudden, it reports that disk was
> reattached. I'm using zfs, and a redundant pool, so, besides the fact
> that it's bad by itself, everything should continue to work, but instead
> my FreeBSD starts to livelock, and all disk/zfs commands stop working -
> they just hang (however, I'm able to log in, launch gstat and other
> utilities, and most of the services seem to be unaffected, however,
> iSCSI target starts to behave weirdly too). I wrote a letter to the LSI
> support, but I guess they will first ask me to try the latest driver,
> which for some reason isn't commited in the tree yet. So, two questions:
>
> - why the latest driver (6.710.11.00) isn't in the main tree, it's from
> december 2015, so at least it seems like it should have been made it to
> the 11.x. In 11.x I still see the 06.709.07.00-fbsd ? may be I don't
> unerstand something.
>
> - has anyone seen anything similar on mrsas ? I'm kinda open to the
> ideas. To be honest, I suspect the controller and the firmware, because
> with factory-default firmware version I had seen all sorts of madness,
> including disk detaching after requesting SMART data with smartmontools.
> This was reproducible and solved with the firmware update, but
> considering the amount of fixes that each new version does contain, I
> suppose there's many of the bugs that left unfixed.
with PERC H330.
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=200993
Miroslav Lachman
Mike Tancsa
Jul 29, 2016, 11:15:16 AM7/29/16
to
On 7/29/2016 9:41 AM, Eugene M. Zheganin wrote:
> Hi.
>
> I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
> Hi.
>
> I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
> - has anyone seen anything similar on mrsas ? I'm kinda open to the
> ideas. To be honest, I suspect the controller and the firmware, because
A somewhat related datapoint as I dont have that exact controller, but I
> ideas. To be honest, I suspect the controller and the firmware, because
am using the driver with a similar class card with very stable results
on RELENG_10. However, smartctl is broken when trying to read from them.
mrsas0@pci0:1:0:0: class=0x010400 card=0x080915d9 chip=0x005d1000
rev=0x02 hdr=0x00
vendor = 'LSI Logic / Symbios Logic'
device = 'MegaRAID SAS-3 3108 [Invader]'
It has a pair of SSD drives that are used in a zfs array for logging and
caching.
0-zoo-# camcontrol devlist
<LSI SMC3108 4.21> at scbus0 target 0 lun 0 (pass0,da0)
<LSI SMC3108 4.21> at scbus0 target 1 lun 0 (pass1,da1)
<WDC WD3003FZEX-00Z4SA0 01.01A01> at scbus2 target 0 lun 0 (ada0,pass2)
<WDC WD3003FZEX-00Z4SA0 01.01A01> at scbus3 target 0 lun 0 (ada1,pass3)
<WDC WD3003FZEX-00Z4SA0 01.01A01> at scbus4 target 0 lun 0 (ada2,pass4)
<AHCI SGPIO Enclosure 1.00 0001> at scbus6 target 0 lun 0 (pass5,ses0)
<WDC WD4003FZEX-00Z4SA0 01.01A01> at scbus7 target 0 lun 0 (ada3,pass6)
<WDC WD4003FZEX-00Z4SA0 01.01A01> at scbus8 target 0 lun 0 (ada4,pass7)
<WDC WD4003FZEX-00Z4SA0 01.01A01> at scbus9 target 0 lun 0 (ada5,pass8)
<WDC WD4003FZEX-00Z4SA0 01.01A01> at scbus10 target 0 lun 0 (ada6,pass9)
<AHCI SGPIO Enclosure 1.00 0001> at scbus13 target 0 lun 0 (pass10,ses1)
0-zoo-#
I am not able to get smartctl to work with them unfortunately.
MegaCli -AdpAllInfo -aAll
Adapter #0
==============================================================================
Versions
================
Product Name : LSI 3108 MegaRAID ROMB
Serial No :
FW Package Build: 24.1.1-0001
Mfg. Data
================
Mfg. Date : 00/00/00
Rework Date : 00/00/00
Revision No :
Battery FRU : N/A
Image Versions in Flash:
================
BIOS Version : 6.13.00_4.14.05.00_0x06010600
Ctrl-R Version : 5.01-0004
FW Version : 4.210.10-2910
NVDATA Version : 3.1310.00-0054
Boot Block Version : 3.00.00.00-0009
--
-------------------
Mike Tancsa, tel +1 519 651 3400
Sentex Communications, mi...@sentex.net
Providing Internet services since 1994 www.sentex.net
Cambridge, Ontario Canada http://www.tancsa.com/
Eugene M. Zheganin
Aug 3, 2016, 6:42:22 AM8/3/16
to
Hi.
On 29.07.2016 20:14, Mike Tancsa wrote:
> On 7/29/2016 9:41 AM, Eugene M. Zheganin wrote:
>> Hi.
>>
>> I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
>> - has anyone seen anything similar on mrsas ? I'm kinda open to the
>> ideas. To be honest, I suspect the controller and the firmware, because
> A somewhat related datapoint as I dont have that exact controller, but I
> am using the driver with a similar class card with very stable results
> on RELENG_10. However, smartctl is broken when trying to read from them.
>
> mrsas0@pci0:1:0:0: class=0x010400 card=0x080915d9 chip=0x005d1000
> rev=0x02 hdr=0x00
> vendor = 'LSI Logic / Symbios Logic'
> device = 'MegaRAID SAS-3 3108 [Invader]'
It's another chip, actually, but thanks for info. Mine is LSISAS3008:
vendor = 'LSI Logic / Symbios Logic'
device = 'MegaRAID SAS-3 3008 [Fury]'
class = mass storage
subclass = RAID
Looks like your problem with SMART may come from a very old firmware, my
controller had same problem - disk was simply detaching after requesting
SMART data. This was fixed in the recent firmware version. I suggest you
should try the firmware upgrade.
Eugene.
On 29.07.2016 20:14, Mike Tancsa wrote:
> On 7/29/2016 9:41 AM, Eugene M. Zheganin wrote:
>> Hi.
>>
>> I'm experiencing some weird troubles with an LSI MegaRAID SAS9341-4i
>> - has anyone seen anything similar on mrsas ? I'm kinda open to the
>> ideas. To be honest, I suspect the controller and the firmware, because
> A somewhat related datapoint as I dont have that exact controller, but I
> am using the driver with a similar class card with very stable results
> on RELENG_10. However, smartctl is broken when trying to read from them.
>
> mrsas0@pci0:1:0:0: class=0x010400 card=0x080915d9 chip=0x005d1000
> rev=0x02 hdr=0x00
> vendor = 'LSI Logic / Symbios Logic'
> device = 'MegaRAID SAS-3 3108 [Invader]'
vendor = 'LSI Logic / Symbios Logic'
class = mass storage
subclass = RAID
controller had same problem - disk was simply detaching after requesting
SMART data. This was fixed in the recent firmware version. I suggest you
should try the firmware upgrade.
Eugene.
Mike Tancsa
Aug 3, 2016, 12:05:35 PM8/3/16
to
On 8/3/2016 6:41 AM, Eugene M. Zheganin wrote:
> Looks like your problem with SMART may come from a very old firmware, my
> controller had same problem - disk was simply detaching after requesting
> SMART data. This was fixed in the recent firmware version. I suggest you
> should try the firmware upgrade.
Thanks! What version are you running ?
> Looks like your problem with SMART may come from a very old firmware, my
> controller had same problem - disk was simply detaching after requesting
> SMART data. This was fixed in the recent firmware version. I suggest you
> should try the firmware upgrade.
---Mike
--
-------------------
Mike Tancsa, tel +1 519 651 3400
Sentex Communications, mi...@sentex.net
Providing Internet services since 1994 www.sentex.net
Cambridge, Ontario Canada http://www.tancsa.com/
0 new messages