I have been battling with the mercury install on my own hosted server (not aws, rackspace or any of the other hosts currently supported).
I have full root access on a clean jaunty installation and I been trying to go through the instructions on http://groups.drupal.org/node/50408
Everything seems to fine up until:
'9) Start the BCFG2 server:
Wait until the bcfg2-server spins up - you know the server is ready when see "serve_forever() [start]".
sudo /etc/init.d/bcfg2-server start; tail -f /var/log/syslog'
I cannot get it to run, I just keep getting errors like:
May 24 12:57:39 s15400895 bcfg2-server[5361]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/universe/binary-amd64/P...
May 24 12:57:55 s15400895 bcfg2-server[5361]: Failed to read file probed.xml
May 24 12:57:55 s15400895 bcfg2-server[5361]: Loading experimental plugin(s): Packages
May 24 12:57:55 s15400895 bcfg2-server[5361]: NOTE: Interfaces subject to change
May 24 12:57:55 s15400895 bcfg2-server[5361]: Failed to bind to socket
May 24 12:57:55 s15400895 bcfg2-server[5361]: Server startup failed
May 24 13:00:01 s15400895 /USR/SBIN/CRON[5366]: (root) CMD ([ -x /usr/sbin/update-motd ] && /usr/sbin/update-motd hourly 2>/dev/null)
May 24 13:00:01 s15400895 /USR/SBIN/CRON[5371]: (root) CMD ([ -x /usr/sbin/update-motd ] && /usr/sbin/update-motd 2>/dev/null)
I can only assume it is because not enough info is provided at step 6 on setting up a group for a non-aws/rackspace installation.
What I have tried:
picking a pre-made group such as:
and swapping out 'aws' and 'amazon-web-services' for '1and1'. (my host).
I also added:
to the bottom of the page but this does not work.
I have even tried using the pre-made amazon setup but this also doesn't work. I also altered the clients.xml to reflect what I had changed in groups.xml.
I would really like to get this up and running asap so any help would be greatly appreciated!!
Comments
If you are able to use Lucid,
If you are able to use Lucid, you may want to give Mercury 1.1 Beta a try. The install instructions can be found here: http://groups.drupal.org/node/70268. For non-specific host - when you get to step #5, use 'sudo bcfg2 -vqed'
Alternatively if that is not an option at the moment, can you post your groups.xml and clients.xml
"Failed to bind to socket"
There's something pretty simple and basic wrong with your bcfg2 config. If you are dying with "Failed to bind to socket" then the server is unable to set itself up to run.
Is it possible you're ALREADY running bcfg2, and trying to start it again?
https://pantheon.io | http://www.chapterthree.com | https://www.outlandishjosh.com
This may be possible. It
This may be possible.
It therefore leads the question "how would i know?" Perhaps more documentation on how to generally maintain this system, starting it, stopping it, monitoring it etc.
As for using Lucid, that's what I would prefer but the host only provides Ubuntu 8.04. I can upgrade it remotely upto and including karmic but if I try to upgrade to Lucid, on reboot there is something in startup which locks down all ports so you can't remote back into it. Any pointers on over-coming this would also be welcome.
I was facing this problem upgrading to the other versions (pre lucid) but found that if i chose to keep the original config settings such as dhclients, grub etc then it would not lock-down on reboot.
I will re-image today, start a fresh install, run it once and see if any errors occur, I will then post the groups and clients xmls.
Thanks
Same build, different issue
I was getting the same errors as the original poster with Jaunty, and after shutting down bcfg2 and restarting I managed to clear some of the errors.
Now, when I kick off the command
sudo bcfg2 -vqI receive the following error message:Server failure: Protocol Error: 401 Unauthorized
Failed to download probes from bcfg2
Server Failure
Any thoughts why I should get authentication errors?
Branden
re: Same build, different issue
This errors sounds to me like the BCFG2 client can't find or authenticate with the BCFG2 server.
In Step 3 (http://groups.drupal.org/node/70268):
The easiest answer to give are "localhost" and "https://localhost:6789" respectively. This way, even if the hostname changes, BCFG2 will still work. I'll add this to the install docs.
Hope this helps,
Greg
--
Greg Coit
Systems Administrator
http://www.chapterthree.com
Im now at the same hurdle as
Im now at the same hurdle as you clockwork robot.
After starting again and running "/etc/init.d/bcfg2-server start; tail -f /var/log/syslog"
I get:
root@s15400895:~# /etc/init.d/bcfg2-server start; tail -f /var/log/syslog
Starting Configuration Management Server: * bcfg2-server
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: Failed to instantiate plugin Pkgmgr
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: Failed to load /var/lib/bcfg2/Rules indices
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: Failed to instantiate plugin Rules
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: Failed to load SSHbase repository from /var/lib/bcfg2/SSHbase
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: (2, 'No such file or directory')
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: Failed to instantiate plugin SSHbase
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: No Metadata Plugin loaded; failed to instantiate Core
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: ('No Metadata Plugin',)
Jun 5 16:34:17 s15400895 bcfg2-server[4744]: exiting
Jun 5 16:40:01 s15400895 /USR/SBIN/CRON[4767]: (root) CMD ([ -x /usr/sbin/update-motd ] && /usr/sbin/update-motd 2>/dev/null)
Jun 5 16:46:25 s15400895 bcfg2-server[4787]: Packages: File read failed; falling back to file download
Jun 5 16:46:25 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/main/binary-amd64/Packa...
Jun 5 16:46:26 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/main/binary-i386/Packag...
Jun 5 16:46:27 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/multiverse/binary-amd64...
Jun 5 16:46:27 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/multiverse/binary-i386/...
Jun 5 16:46:27 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/restricted/binary-amd64...
Jun 5 16:46:27 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/restricted/binary-i386/...
Jun 5 16:46:28 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/universe/binary-amd64/P...
Jun 5 16:46:31 s15400895 bcfg2-server[4787]: Packages: Updating http://us.archive.ubuntu.com/ubuntu/dists/jaunty/universe/binary-i386/Pa...
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: Failed to read file probed.xml
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: Loading experimental plugin(s): Packages
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: NOTE: Interfaces subject to change
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: service available at https://s15400895:6789
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: serving bcfg2-server at https://s15400895:6789
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: serve_forever() [start]
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: Handled 35 events in 0.012s
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: Handled 31 events in 0.002s
Jun 5 16:46:54 s15400895 bcfg2-server[4787]: Handled 83 events in 0.030s
Jun 5 16:46:55 s15400895 bcfg2-server[4787]: Handled 64 events in 0.023s
Jun 5 16:46:55 s15400895 bcfg2-server[4787]: Handled 38 events in 0.022s
Jun 5 16:46:55 s15400895 bcfg2-server[4787]: Handled 7 events in 0.005s
and it says 'serve_forever()' which it never before, albeit with a whole stack of errors.
If I then try "bcfg2 -vq"
I get :
Server failure: Protocol Error: 401 Unauthorized
Failed to download probes from bcfg2
Server Failure
The difference this time was using 'localhost' and 'http://localhost:6789' during the setup
Resolved
I fixed my 401 errors by looking at the hostnames that were being specified to the bfcg2 server via the configuration vs what my system reported. So in my case, all of my config files (bcfg2.conf...) were looking for a server with credentials at http://tatiana:6789 while my host was resolving itself to be tatiana.example.org due to my /etc/hosts file setup. When I got the two hosts to match, I stopped getting 401 issues.
In the end, however, I had a pretty ropey setup on jaunty due to missing out some steps from the instructions. I did a rebuild of the whole project on the new Lucid architecture with the new instruction set ( http://groups.drupal.org/node/70268 ) and it worked a treat. If you can move to Lucid, then this is the simplest way to make progress.
Good luck
Unfortunatly the vps I have
Unfortunatly the vps I have is not able to upgrade to Lucid, incomaptible kernals apparently. I have found a different host that I am looking to move to, called webcurl. They specialise in hosting drupal sites and all sites are setup using Mercury so having them manage the Mercury stack will take the responsibility away from me :)
Until then id still like to get this running for myself even if only for a while until I migrate.
It does appear that Im having the same host resolve issue. What exactly did you change in /etc/hosts?
thanks for this
Months later your flagging of the hostname issue helps a ton :)