Hi all,
I have a BladeSystem c7000 Enclosure with 16 blades.
They are all running RHEL4 Update 8.
The blades are connected to a tape storage unit over fibre.
When I reboot one of the blades, the boot process hangs at the initializing hardware step. Closer investigation (using strace) reveals it is during the running of /sbin/kmodule and more specifically when it probes /dev/nst2 The output from strace looks like this
11:21:19 access("/dev/nst1", F_OK) = 0
11:21:19 open("/dev,nst1",O_RDONLY|O_NONBLOCK)= 3
11:21:19 ioctl(3, CDROMAUDIOBUFSIZ or SCSI_IOCTL_GET_IDLUN, 0x7fbfffab58) = 0
11:21:19 close(3) = 0
11:21:19 access("/dev/nst2", F_OK) = 0
11:21:19 open("/dev/nst2",O_RDONLY|O_NONBLOCK) = -1 EIO (Input/output error)
13:11:22 access("/dev/nst3", F_OK) = 0
13:11:22 open("/dev/nst3",O_RDONLY|O_NONBLOCK) = 3
13:11:22 ioctl(3, CDROMAUDIOBUFSIZ or SCSI_IOCTL_GET_IDLUN, 0x7fbfffab58) = 0
13:11:22 close(3) = 0
You can see the open("/dev/nst2" line ends in EIO (Input/output error) after nearly 2 hours.
The servers were successfully rebooted last time on Feb 7th. If I remove the tape device from the servers, they boot fine. I have tried updatig the Proliant Support pack, and HBA firmware on one of the servers, but this didn't make a difference.
How can I find out more about what has changed, and what is causing this issue ?
Thanks
