2

I have a server running RHEL 5.5 with EMC PowerPath installed. Due to the security requirement, I had to upgrade the kernel to a newer version. Unfortunately, I was not aware of the impact of the upgrade on EMC PowerPath. After the upgrade, the server cannot load the PowerPath kernel modules anymore.

So, I downgraded back to the original kernel. Still the kernel modules refused to load.

I tried up load the modules manually, but get the following error.

[root@cvoapd05 EMCpower.LINUX-5.3.1.02.00]# modprobe emcp
FATAL: Error inserting emcp (/lib/modules/2.6.18-194.32.1.el5/powerpath/emcp.ko): Device or resource busy

The .ko files are still in the path "/lib/modules/2.6.18-194.32.1.el5/powerpath", and the file sizes and dates are unchanged.

Is there anyway I can recover from this problem?

grub.conf:

# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE:  You have a /boot partition.  This means that
#          all kernel and initrd paths are relative to /boot/, eg.
#          root (hd0,0)
#          kernel /vmlinuz-version ro root=/dev/cciss/c0d0p8
#          initrd /initrd-version.img
#boot=/dev/cciss/c0d0
# default=0 # commented out by Proliant HBA install script
default=0
fallback=1
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
hiddenmenu
title Red Hat Enterprise Linux Server (2.6.18-194.32.1.el5)
        root (hd0,0)
        kernel /vmlinuz-2.6.18-194.32.1.el5 ro root=LABEL=/ rhgb quiet crashkernel=1024M@16M
        initrd /initrd-2.6.18-194.32.1.el5.img

title Red Hat Enterprise Linux Server (2.6.18-194.el5)
        root (hd0,0)
        kernel /vmlinuz-2.6.18-194.el5 ro root=LABEL=/ rhgb quiet crashkernel=1024M@16M
        initrd /initrd-2.6.18-194.el5.img

# This entry (no. 1) added by Proliant HBA install script
# in package cpq_cciss-3.6.26-5.rhel5
title HP-2.6.18-194.el5
        root (hd0,0)
        kernel /vmlinuz-2.6.18-194.el5 ro root=LABEL=/ rhgb quiet crashkernel=1024M@16M
        initrd /HP-initrd-2.6.18-194.el5.img
kjloh
  • 117
  • 2
  • 4
  • 7
  • Post the output of uname -a and the contents of your /etc/grub.conf – dmourati May 26 '11 at 04:54
  • Uname as follows: `[root@cvoapd05 ~]# uname -a Linux cvoapd05 2.6.18-194.32.1.el5 #1 SMP Mon Dec 20 10:52:42 EST 2010 x86_64 x86_64 x86_64 GNU/Linux` Grub.conf is added to the question above. – kjloh May 26 '11 at 06:19
  • Does anything show up in `dmesg` when you try loading the module? That's typically where load-time errors show up. –  May 29 '11 at 00:22
  • Take a look at `lsmod | grep emcp` – quanta May 28 '11 at 17:36

3 Answers3

2

There is probably another driver attached to a device that the powerpath driver is trying to attach to. Check to see if there are any other multipath drivers loaded, lsmod | grep path should be sufficient. In fact, simply providing the output of lsmod should help us determine what is causing this issue.

Joe
  • 58
  • 5
1

Use the native multipath-daemon instead of powerpath.

Nils
  • 7,695
  • 3
  • 34
  • 73
  • Do you have any evidence that one is better than the other? – Mark Wagner Aug 05 '11 at 18:38
  • 1
    It is alwas bad to have vendor specific special changes in your os - you currently experience that. I found the same to be true with attaching linux to a HP-EVA-SAN (with SLES10 I need a special HBA-driver from HP, needs to be recompiled at every kernel-update). Take a look at multipathd - there is almost certainly emc-powerpath-support already built in. – Nils Aug 05 '11 at 19:23
  • With Linux native multipath i never found this problem – c4f4t0r Sep 19 '13 at 17:44
  • @Mark see above comments. Give it a try... – Nils Sep 19 '13 at 21:18
1

/etc/rc.d/PowerPath has logic in it to determine if a different kernel has been loaded and if so update its configuration, which involves more than copying emcp.ko.

Try manually running /etc/rc.d/PowerPath start to either fix the problem or get more information on what the problem is.

Mark Wagner
  • 18,019
  • 2
  • 32
  • 47