Dne 18.5.2018 v 17:40 Casey & Gina napsal(a):
Thank you - I hadn't seen the "releases" link on github before and somehow
missed that. Sorry for that. I thought there would be download links somewhere on the
clusterlabs website. I will try compiling this today to try.
I finally managed to get pcs-0.9.164 compiled and installed. There were a
number of dependencies to figure out, so I thought I'd send what I ended up
doing in case it helps somebody else (using Ubuntu 16.04). I needed to do the
following:
------
apt install python3-setuptools python3-lxml fonts-liberation fontconfig
libpam0g-dev bundler ruby-rpam-ruby19
------
Those are all listed in pcs readme except ruby rpam which is downloaded
automatically during pcs installation.
I'm not sure that the ruby-rpam-ruby19 package installation is actually
necessary, as it didn't seem to work:
------
Could not find rpam-ruby19-1.2.1.gem for installation
------
So, I did this as well, which I guess did the same thing the package install
should have done?:
------
gem install rpam-ruby19 -v '1.2.1'
------
Anyways, after that I started up the pcsd service, and proceeded to try to
start the cluster. This appeared to work okay, until I tried to do `pcs
cluster start --all`, which resulted in HTTP 400 errors. Using --debug, I saw
that it was failing to start corosync because the directory /var/log/cluster
did not exist. Not sure why this would be, since corosync was installed using
the apt package, and with the other pcs version it used /var/log/corosync as
the logging directory. Nonetheless I created /var/log/cluster, and then the
cluster started okay.
That's what pcsd/settings.rb.debian and pcs/settings.py.debian are for.
Those files contain proper paths for Debian-based distros. Maybe the log
path is missing in one of them? Can you post the output with --debug?
Finally, I tried adding the stonith resource with the following command, and
ran into an error that I don't know how to solve:
------
pcs stonith create vfencing external/vcenter VI_SERVER=10.124.137.100
VI_CREDSTORE=/etc/pacemaker/vicredentials.xml
HOSTLIST="d-gp2-dbpg0-1=d-gp2-dbpg0-1;d-gp2-dbpg0-2=d-gp2-dbpg0-2;d-gp2-dbpg0-3=d-gp2-dbpg0-3"
RESETPOWERON=0
Error: unable to run command /usr/lib/x86_64-linux-gnu/pacemaker/stonithd
metadata: No such file or directory:
'/usr/lib/x86_64-linux-gnu/pacemaker/stonithd'
------
I'm not sure why it's including x86_64-linux-gnu in the path - the actual path,
as provided by the pacemaker package, is /usr/lib/pacemaker/stonithd. Is there
a way to set this path correct when compiling/installing pcs from the source
package?
This is how it's configured for Debian-based distros in
pcs/settings.py.debian
https://github.com/ClusterLabs/pcs/commit/05192c3c02acb7ac28a8c02022725451546e499f
Pcs development team don't run Debian nor Ubuntu so we cannot really
tell if it's correct or not.
Have you tried installing pcs package from newer Ubuntu? It might be
easier...
Regards,
Tomas
Regards,
_______________________________________________
Users mailing list: [email protected]
https://lists.clusterlabs.org/mailman/listinfo/users
Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org
