EVA3000 linux LVM problem

Occasional Contributor

EVA3000 linux LVM problem

in our setup we have hpdl320 installed with RHEL4ES and a scsi stroage is attached storage is accessed thru LVM. as the day passed we run out of disk space adn decided to
add configure EVA3000 with linux

As a part of procedure we have added Emulex lite pulse HBA and done zoning a created a vdisk of 32gb each visible from server

The problem iam facing are follows

once if i create a pv,vg and lv and reboot
the server system goes to maintanence mode saying superblock of pv is corupted and tells to run fsck on second super block i done it but no resonse then i commented the vg entry in vfstab and booted the system .

Now i created a normal 3xt3 filesystem and created mount pointed entered in /etc/vfstab and rebooted same problem system going to maintance mode and another problem is device number has changed

that means the mount point which i access data from /dev/sdb1 is now /dev/sdn1

ANy one who come across this problem plz help

to the frank i must say we have not installed a EMULEX driver since we dont have a down time and its very critical server
Steven E. Protter
Exalted Contributor

Re: EVA3000 linux LVM problem


You may need to build a new ram disk and build the server.

modprobe.conf changes:

alias scsi_hostadapter1 lpfc
options lpfc lpfc_nodev_tmo=14 lpfc_lun_queue_depth=30 lpfc_discovery_threads=1

cd /boot
mv initrd-....img to initrd-....img.orig


mkinitrd initrd-2.6.9-42.0.3.ELsmp.img 2.6.9-42.0.3.ELsmp

Obviously you will have to use a kernel image that is relavent to your system.

Steven E Protter
Owner of ISN Corporation
Sponsor: http://hpux.ws
Twitter: http://twitter.com/hpuxlinux
Founder http://newdatacloud.com