Hello all, I’m at my wits end trying to resolve this. We are running centos 4.5 on Intel hardware. Dual SCSI disk drives mirrored on an LSI Logic controller. Every once in a while and not always on the same server and not only on the local SCSI Drives. System A – Dual internal drives on /dev/sda System B – Dual internal drives on /dev/sdc with a DAS on /dev/sda. Each of these systems experienced a kernel mptbase error and placed /dev/sda into read only mode. Note again the /dev/sda isn’t always local. For system A – remounting in ro mode didn’t work and the system had to be rebooted. File system check and bad block checks showed nothing and when the system was rebooted – it was fine. A portion of the messages log is below. I would appreciate any ideas or directions. Thanks, Steve Moccio Aug 7 01:00:06 sshd(pam_unix)[18336]: session opened for user root by (uid=0) Aug 7 09:00:36 kernel: mptscsi: ioc1: attempting task abort! (sc=f6f07c80) Aug 7 09:00:36 kernel: scsi1 : destination target 0, lun 0 Aug 7 09:00:36 kernel: command = Write (10) 00 00 00 fb d7 00 01 90 00 Aug 7 09:00:38 kernel: mptbase: Initiating ioc1 recovery Aug 7 09:00:44 kernel: drivers/message/fusion/mptctl.c at 1985::mptctl_do_mpt_command - Busy with IOC Reset Aug 7 09:01:19 last message repeated 10 times Aug 7 09:01:40 last message repeated 7 times Aug 7 09:01:41 kernel: mptbase: ioc1: ERROR - Diagnostic reset FAILED! (102h) Aug 7 09:01:41 kernel: mptbase: ioc1 NOT READY WARNING! Aug 7 09:01:41 kernel: mptbase: WARNING - (-1) Cannot recover ioc1 Aug 7 09:01:41 kernel: mptscsi: ioc1: Issue of TaskMgmt failed! Aug 7 09:01:41 kernel: mptscsi: ioc1: task abort: FAILED (sc=f6f07c80) Aug 7 09:01:41 kernel: mptscsi: ioc1: attempting bus reset! (sc=f6f07c80) Aug 7 09:01:41 kernel: scsi1 : destination target 0, lun 0 Aug 7 09:01:41 kernel: command = Write (10) 00 00 00 fb d7 00 01 90 00 Aug 7 09:01:41 kernel: mptbase: Initiating ioc1 recovery Aug 7 09:01:46 kernel: mptbase: ioc1: ERROR - Doorbell ACK timeout (count=4999), IntStatus=80000000! Aug 7 09:01:47 kernel: drivers/message/fusion/mptctl.c at 1985::mptctl_do_mpt_command - Busy with IOC Reset Aug 7 09:02:23 last message repeated 10 times Aug 7 09:02:44 last message repeated 7 times Aug 7 09:02:47 kernel: mptbase: ioc1: ERROR - Diagnostic reset FAILED! (102h) Aug 7 09:02:47 kernel: mptbase: ioc1 NOT READY WARNING! Aug 7 09:02:47 kernel: mptbase: WARNING - (-1) Cannot recover ioc1 Aug 7 09:02:47 kernel: mptscsi: ioc1: bus reset: FAILED (sc=f6f07c80) Aug 7 09:02:48 kernel: mptscsi: ioc1: Attempting host reset! (sc=f6f07c80) Aug 7 09:02:48 kernel: mptbase: Initiating ioc1 recovery Aug 7 09:02:51 kernel: drivers/message/fusion/mptctl.c at 1985::mptctl_do_mpt_command - Busy with IOC Reset Aug 7 09:02:51 kernel: drivers/message/fusion/mptctl.c at 1985::mptctl_do_mpt_command - Busy with IOC Reset Aug 7 09:02:53 kernel: mptbase: ioc1: ERROR - Doorbell ACK timeout (count=4999), IntStatus=80000000! Aug 7 09:02:58 kernel: drivers/message/fusion/mptctl.c at 1985::mptctl_do_mpt_command - Busy with IOC Reset Aug 7 09:03:34 last message repeated 10 times Aug 7 09:03:48 last message repeated 5 times Aug 7 09:03:54 kernel: mptbase: ioc1: ERROR - Diagnostic reset FAILED! (102h) Aug 7 09:03:54 kernel: mptbase: ioc1 NOT READY WARNING! Aug 7 09:03:54 kernel: mptbase: WARNING - (-1) Cannot recover ioc1 Aug 7 09:03:54 kernel: scsi: Device offlined - not ready after error recovery: host 1 channel 0 id 0 lun 0 -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.centos.org/pipermail/centos/attachments/20080827/b73eb9a2/attachment-0004.html>