![]() |
SCSI-3 Persist Reservation using cciss driver
I have a Compaq Proliant DL380 server with Redhat Enterprise Linux 3.0 ES at work.
We need to be able to do a SCSI-3 Persist Reservation,using sg_persist command. It is included in sg3_utils pakage. We need to lock sharing disks connected by SCSI for cluster system. Here's the problem. In order to use sg_persist command, sg device is required. it needs to link to SCSI storage devices. But now cciss devices works as SCSI device. In /proc/scsi/scsi, nothing appears. I tried not to use cciss devices but to use DL380. But It seemed DL380 requires cciss modules. sg module is installed, (appears after lsmod), and there are /dev/sg* devices,but it is not used. Does anyone know how i can make this? |
Additional Infomation.
Additional Infomation.
# cat /proc/scsi/scsi Attached devices: none # dmesg | grep cciss cciss: Device 0x46 has been found at bus 4 dev 3 func 0 cciss/c0d0: p1 p2 cciss/c0d1: p1 EXT3 FS 2.4-0.9.19, 19 August 2002 on cciss0(104,1), internal journal #dmesg | grep SCSI SCSI subsystem driver Revision: 1.00 # cat /etc/modules.conf | grep cciss alias scsi_hostadapter cciss # lsmod Module Size Used by Not tainted sg 36204 0 (autoclean) (unused) cciss 44036 2 sd_mod 13744 0 (unused) scsi_mod 106408 2 [sg cciss sd_mod] |
Dear,
Did you fix this problem? I am not sure but I think it is the same problem that I have... We have 2 servers Proliant DL380 G5 connected at Storage MSA500 G2. We just install the Cluster in this servers and when we make some tests like put off the Network cable from the one of server the Cluster didn't reach to pass the service to another Node (server). We see in the /etc/messages when the cluster try mount the fileSystem it didn't reach because it request to run "fsck" manually. Thanks. Regards. Ricardo Olguin. Quote:
|
| All times are GMT -5. The time now is 05:06 AM. |