problem with LSI Logic / Symbios Logic 53c1030 PCI-X Fusion-MPT Dual Ultra320 SCSI and a huge raid device

[Eric Doutreleau]

Eric Doutreleau a écrit :
> i have a server with a LSI Logic / Symbios Logic 53c1030 PCI-X 
> Fusion-MPT Dual Ultra320 SCSI and a huge raid device
>
> my server is under centos5.2 kernel 2.6.18-92.1.18.el5 x86_64
>
> on the raid device i have a 7To partition ands a 2To one.
>
> When i try to write a big file ( several dozen of gigabytes ) my 
> serveur freeze.
>
> i got the following message on my console
>
> d 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff81018f3ace00, mf = ffff81022e1873c0, idx=ca
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff81019abde680, mf = ffff81022e187420, idx=cb
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff810227c74380, mf = ffff81022e187660, idx=d1
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff810227c74b00, mf = ffff81022e187780, idx=d4
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff81018f3acc80, mf = ffff81022e187900, idx=d8
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff81020d7bc3c0, mf = ffff81022e187960, idx=d9
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff810211a7ce40, mf = ffff81022e1882c0, idx=f2
> sd 1:0:0:1: mptscsih: ioc0: completing cmds: fw_channel 0, fw_id 0, 
> sc=ffff81022e7cc9c0, mf = ffff81022e1886e0, idx=fd
> mptscsih: ioc0: Issue of TaskMgmt failed!
> mptscsih: ioc0: task abort: FAILED (sc=ffff81017d265200)
> mptscsih: ioc0: attempting task abort! (sc=ffff81018f3ac500)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 62 8a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81018f3ac500)
> mptscsih: ioc0: attempting task abort! (sc=ffff81017d265500)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 6a 8a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81017d265500)
> mptscsih: ioc0: attempting task abort! (sc=ffff81011bf993c0)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 6e 8a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81011bf993c0)
> mptscsih: ioc0: attempting task abort! (sc=ffff81017d265380)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 72 8a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81017d265380)
> mptscsih: ioc0: attempting task abort! (sc=ffff81011bf99540)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 76 8a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81011bf99540)
> mptscsih: ioc0: attempting task abort! (sc=ffff8101416353c0)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 7a 8a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff8101416353c0)
> mptscsih: ioc0: attempting task abort! (sc=ffff8101aff019c0)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 7e 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff8101aff019c0)
> mptscsih: ioc0: attempting task abort! (sc=ffff81022e7cc0c0)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 82 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81022e7cc0c0)
> mptscsih: ioc0: attempting task abort! (sc=ffff810211a7c9c0)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 86 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff810211a7c9c0)
> mptscsih: ioc0: attempting task abort! (sc=ffff810227c74200)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 8a 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff810227c74200)
> mptscsih: ioc0: attempting task abort! (sc=ffff81011bf99e40)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 8e 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81011bf99e40)
> mptscsih: ioc0: attempting task abort! (sc=ffff8101416359c0)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 92 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff8101416359c0)
> mptscsih: ioc0: attempting task abort! (sc=ffff81017d265b00)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 96 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81017d265b00)
> mptscsih: ioc0: attempting task abort! (sc=ffff81018f3ace00)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 9a 92 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81018f3ace00)
> mptscsih: ioc0: attempting task abort! (sc=ffff81018f3ac200)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 9e 9a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81018f3ac200)
> mptscsih: ioc0: attempting task abort! (sc=ffff81018f3ac980)
> sd 1:0:0:1:
> command: Write(10): 2a 00 01 73 a2 9a 00 04 00 00
> mptscsih: ioc0: task abort: SUCCESS (sc=ffff81018f3ac980)
> mptscsih: ioc0: attempting task abort! (sc=ffff810211a7c3c0)
> sd 1:0:0:1:
>
> When i try to do that on the 2To partition i have no problem.
>
> on the Raid device all is ok
> i got no alert about some problem it could have.
>
> When i was on 5.1 system i didn't have that kind of problem
>
> I installed the latest lsi driver.
> It seems to ameliorate a bit the situation.
> the scsi interface hung a bit later
> i got the messages
> mptbase: ioc0: WARNING - IOC is in FAULT state (000eh)!!!
> mptbase: ioc0: WARNING - Issuing HardReset from mpt_fault_reset_work!!
> mptbase: ioc0: Initiating recovery
> mptbase: ioc0: WARNING - IOC is in FAULT state!!!
>           FAULT code = 000eh
> mptbase: ioc0: ERROR - Doorbell ACK timeout (count=14999), 
> IntStatus=80000001!
> mptbase: ioc0: Recovered from IOC FAULT
> mptbase: ioc0: WARNING - mpt_fault_reset_work: HardReset: success
> target1:0:0: Beginning Domain Validation
> sd 1:0:0:1: SCSI error: return code = 0x000b0000
> end_request: I/O error, dev sdb, sector 9475660634
> Buffer I/O error on device sdb1, logical block 1184457575
> lost page write due to I/O error on sdb1
> sd 1:0:0:1: SCSI error: return code = 0x000b0000
> end_request: I/O error, dev sdb, sector 9475637026
> Buffer I/O error on device sdb1, logical block 1184454624
> lost page write due to I/O error on sdb1
> sd 1:0:0:1: SCSI error: return code = 0x000b0000
> end_request: I/O error, dev sdb, sector 9475633274
>
> does someone knows any special issue with lsi logic and centos 5.2?
>
> thanks in advance for any help
>
the problem disapeared when i downgrade the kernel to the 5.1 version.
but it s not a long term solution