If you attach to linux box a disk partitioned under FreeBSD or OS X, you will find this warning message on fdisk command:
"WARNING: GPT (GUID Partition Table) detected on '/dev/sdc'! The util fdisk doesn't support GPT. Use GNU Parted."
During mksysb process you have the chance to encounter many errors. One of them is "An internal packing error occurred: incorrect size" and I will try below to aid you to get pass it.
# mksysb -eiX /backups-tmp/gzlpar-rootvg.sysb
Creating information file (/image.data) for rootvg.
Creating list of files to back up
Backing up 89961 files......................[color=red]An internal packing error occurred: incorrect size
./var/adm/wtmp will not be backed up.[/color]
Debian is great but sometimes basic simple things are missing so we have to step in.
For example, Debian is coming by default with /etc/rc.local where you can put your custom commands which are automatically executed on startup. But what about custom stop script to be executed on shutdown or reboot? This is missing.
No problem! All you can do is follow the instructions below.
1. Edit or replace /etc/init.d/rc.local with this one:
root@pluto:~# cat /etc/init.d/rc.local
### BEGIN INIT INFO
# Provides: rc.local
Before starting, I have to mention this tutorial is not about ZFS Fuse which comes by default with Debian. This is the real deal, the native ZFS for Linux, without fuse support. In theory should be faster and more reliable and according to my tests so far, it is amazing!
First of all, install ZFS:
UPDATE: ZFS packages are included in debian repository. So all you have to do now is:
How to move the services manually, to the secondary node (candy):
This will also prevent the resource to automatically failover if the primary node is back online so keep that in mind when you plan several reboots on primary node and you don't want to be bothered. crm resource migrate nfs-group candy
STONITH is short for "Shoot The Other Node In The Head". It's a technique for fencing in computer clusters.
Fencing is the isolation of a failed node so that it does not cause disruption to a computer cluster. As its name suggests, STONITH fences failed nodes by resetting or powering down the failed node.
Multi-node error-prone contention in a cluster can have catastrophic results, such as if both nodes try writing to a shared storage resource. STONITH provides effective, if rather drastic, protection against these problems.
The examples below are done on an environment where drbd volume resides on lvm logical volume and the filesystem is reiserfs.
a. If you want to shrink it, you will have to reduce first the filesystem on the primary node then the logical volume on both nodes. Afterwards you execute the resize command for drbd on primary node.
Be careful, in order to shrink the filesystem, you will have to umount it and obviously to stop the services before that (nfs, apache, etc.).
In this situation, ssh was unable to start because the wpar was restored from standalone machine backup (mksysb).
However, even if this is not your case, you can try to see if the procedure will fix your ssh.
SSH doesn't want to start:
gzaix@root:/home/root # lssrc -s sshd
Subsystem Group PID Status
sshd ssh inoperative
gzaix@root:/home/root # startsrc -s sshd
0513-059 The sshd Subsystem has been started. Subsystem PID is 18415656.