>
> Under load, I'm seeing occasional controller resets, and then some i/o
>> timeouts on disks owned by the controller being reset.
>>
>
> What kind of timeouts are you seeing? Are they on the initiator and if so
> can you send the /var/log/messages? If there are nop/ping iscsi timeouts
> then it may be a bug, where open-iscsi was too agressive in determining if
> there was a timeout.
>

On the servers, nothing other than no access to storage for 30-60 seconds.
But the MD3000i logs this - which seems to be a controller resetting itself,
the interesting parts being the Physical Disk path redundancy lost which
should be related to the opensource rdac driver, according to Dell. OS
connected is Debian, which Dell for some reason chose not to support ;-)

*10-03-16 10:43:31** ** Physical Disk** **Enclosure 0, Slot 14** **Physical
Disk path redundancy restored*
*10-03-16 10:43:31** ** Physical Disk** **Enclosure 0, Slot 13** **Physical
Disk path redundancy restored*
*10-03-16 10:43:31** ** Physical Disk** **Enclosure 0, Slot 12** **Physical
Disk path redundancy restored*
10-03-16 10:43:24  Controller Module RAID Controller Module in slot 0 Alternate
RAID controller module checked in late
10-03-16 10:43:10  Sensor Enclosure 0, Slot 1 Temperature changed to optimal
10-03-16 10:42:42  Controller Module RAID Controller Module in slot 0 All
channel reset detected
10-03-16 10:42:42  Controller Module RAID Controller Module in slot 0 AEN
posted for recently logged event
10-03-16 10:42:36  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 1 All connections established through wide port
10-03-16 10:42:36  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 1 Single connection established through previously failed wide port
10-03-16 10:42:36  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 All connections established through wide port
10-03-16 10:42:36  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Single connection established through previously failed wide port
10-03-16 10:43:40  Controller Module RAID Controller Module in slot 1
Start-of-day
routine completed
10-03-16 10:43:29  Controller Module RAID Controller Module in slot 1 Cache
mirroring on RAID controller modules not synchronized
10-03-16 10:43:25  Initiator Host-side: RAID controller module in slot 0,
port - ISCSI carrier has been detected
10-03-16 10:43:24  Initiator Host-side: RAID controller module in slot 0,
port - ISCSI carrier has been detected
10-03-16 10:43:21  Target iqn.2000-04.com.qlogic:qla4052c.fs10515a02997.1 iSCSI
interface restarted
10-03-16 10:43:17  Pack Enclosure 0 RAID Controller Module cache battery is
fully charged
10-03-16 10:43:17  Controller Module Firmware None Premium feature enabled
10-03-16 10:43:17  Controller Module Firmware None Premium feature enabled
10-03-16 10:43:17  Controller Module Firmware None Premium feature enabled
10-03-16 10:43:17  Controller Module Firmware None Premium feature enabled
10-03-16 10:43:15  Controller Module RAID Controller Module in slot 1 RAID
Controller Module reset
10-03-16 10:43:04  Controller Module RAID Controller Module in slot 1
Start-of-day
routine begun
*10-03-16 10:42:33** ** Disk** **Enclosure 0, Slot 14** **Physical Disk path
redundancy lost*
*10-03-16 10:42:33** ** Disk** **Enclosure 0, Slot 13** **Physical Disk path
redundancy lost*
*10-03-16 10:42:33** ** Disk** **Enclosure 0, Slot 12** **Physical Disk path
redundancy lost*
10-03-16 10:42:33  Controller Module RAID Controller Module in slot 0 AEN
posted for recently logged event
10-03-16 10:42:20  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 All connections established through wide port
10-03-16 10:42:19  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Single connection established through previously failed wide port
10-03-16 10:42:02  Controller Module RAID Controller Module in slot 0 Mode
select for redundant RAID controller module page 2C received
10-03-16 10:42:02  Controller Module RAID Controller Module in slot 1 RAID
Controller Module placed online
10-03-16 10:42:01  Controller Module RAID Controller Module in slot 0 Unwritten
data/consistency recovered from cache
10-03-16 10:42:01  Controller Module RAID Controller Module in slot 0 Unwritten
data/consistency recovered from cache
10-03-16 10:42:01  Controller Module RAID Controller Module in slot 0 Unwritten
data/consistency recovered from cache
10-03-16 10:41:59  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Degraded wide port becomes failed
10-03-16 10:41:59  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Optimal wide port becomes degraded
10-03-16 10:41:59  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 All connections established through wide port
10-03-16 10:41:59  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Single connection established through previously failed wide port
10-03-16 10:41:59  Controller Module RAID Controller Module in slot 0 Unwritten
data/consistency recovered from cache
10-03-16 10:41:58  Controller Module RAID Controller Module in slot 0 RAID
Controller Module reset by its alternate
10-03-16 10:41:58  Controller Module RAID Controller Module in slot 1 RAID
Controller Module placed offline
10-03-16 10:41:56  Controller Module RAID Controller Module in slot 0 RAID
Controller Module cache not enabled or was internally disabled
10-03-16 10:41:56  Controller Module RAID Controller Module in slot 0 Cache
mirroring on RAID controller modules not synchronized
10-03-16 10:41:50  Disk None Destination driver error
10-03-16 10:41:43  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Degraded wide port becomes failed
10-03-16 10:41:43  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Optimal wide port becomes degraded
10-03-16 10:41:43  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Degraded wide port becomes failed
10-03-16 10:41:42  Component (EMM, GBIC/SFP, Power Supply, or Fan) Enclosure
0, Slot 0 Optimal wide port becomes degraded
10-03-16 10:41:41  Controller Module RAID Controller Module in slot 0 RAID
Controller Module reset by its alternate
10-03-16 10:41:41  Disk None Destination driver error
10-03-16 10:41:24  Disk None Destination driver error
10-03-16 10:40:51  Disk None Destination driver error


>  Now I'm told by Dell support that the rdac module on their support cd
>> is modified specifically for the md3000i, which is why I'm
>> experiencing these problems.
>>
>>
> is the rdac module scsi_dh_rdac or dm-rdac or what is the name of the
> module?
>

Module name is dm_rdac - backported from 2.6.22 to run on debian etch 2.6.18
according to
http://www.performancemagic.com/Dell1950_MD3000i_Xen_Debian_iSCSI_RDAC/Multipathing.html
Openiscsi is open-iscsi-2.0-865.15

Ideas ?

Thanks

-- 
You received this message because you are subscribed to the Google Groups 
"open-iscsi" group.
To post to this group, send email to open-is...@googlegroups.com.
To unsubscribe from this group, send email to 
open-iscsi+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/open-iscsi?hl=en.

Reply via email to