On 11/23/05, John David <john@nationalbfp.com> wrote:
As a follow-up to my previous post, I have determined that Fedora Core 2
refuses to start the drbd service on Node2.

Why is that, and what can I do to correct it?  Specifically, "service
drbd start" returns an error about " rc.nodeinfo".  This file is present,
and when I manually add the line "drbd   all   Y", I can then issue the
"onnode 2 service drbd start", (which then attempts to start the service
on *every* node!)  But when it did start on Node2, it began to sync with
Node1!  Yay!

So, here are my final hurdles:

1) How can I make Fedora start drbd services automatically at boot?

drbd in the kernel bootparam starts DRBD at bootup, but this doesn't work for netboot unless you tweak ssi-ksync-network.  This is because the startup is done at the very early stages of boot.  Also part of your problems is because you haven't setup a boot partition on node2 so that node2 can reliably boot up the cluster by itself when node1 breaks.

2) How can I make a service, like httpd, start on a different node (say,
node 3)?  Is that necessary, or will bash_ll handle this for me?

man chkconfig

3) If node 3 dies, or gets abducted by aliens, will node 2 fire httpd
back up automatically, since the process fell off?

You need to get the hang of OpenSSI and its userspace tools.  Try the OpenSSI docs.

You said in our wiki that DRBD crashed your cluster?  It's possible (and easy) to setup DRBD in such a way that will DoS the ICS network on which OpenSSI cluster membership heavily relies.  If DRBD is on the ICS, then you need QoS.  Otherwise lower your sync rate and/or place DRBD on a separate physical network.

Also it's been known among the DRBD circle that drbd-0.7 is really meant for kernel-2.6 whereas drbd-0.6 would be more suitable for kernel-2.4.  Since no one ported drbd-0.6 to kernel-2.4, maybe you can try our SSI-1.9.2 preview release for Fedora Core 3 (based on kernel-2.6).