AMI issue with mercury-pressflow-drupal-lucid-64-1.1-Beta2

Bug #607469 reported by Robert Brown (rjbrown99)
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
DEPRECATED Pantheon
Invalid
Undecided
Unassigned

Bug Description

I fired up AMI ID ami-554aa33c today on a large instance, which is mercury-pressflow-drupal-lucid-64-1.1-Beta2. It never seemed to finish booting so I manually ran a 'sudo sh -x /etc/mercury/init.sh' and here's what I got. Looks like there are some bzr issues with pulling down the latest code which causes cascading problems with the instance.

ubuntu@domU-12-31-39-0C-C8-B1:/etc/mercury$ sudo sh -x init.sh
+ [ -e /etc/mercury/incep ]
+
+ cd /var/www/
+ bzr merge --force
+ exec
Merging from remembered parent location http://bazaar.launchpad.net/~pressflow/pressflow/6/
bzr: ERROR: exceptions.AssertionError: Unknown kind 'absent'

*** Bazaar has encountered an internal error. This probably indicates a
    bug in Bazaar. You can help us fix it by filing a bug report at
        https://bugs.launchpad.net/bzr/+filebug
    attaching the crash file
        /home/ubuntu/.cache/crash/bzr-20100719210833-13580.crash
    and including a description of the problem.

    The crash file is plain text and you can inspect or edit it to remove
    private information.
+ cd /var/lib/bcfg2
+ bzr merge --force
Merging from remembered parent location http://bazaar.launchpad.net/~pantheon-developers/pantheon/1.1/
bzr: ERROR: Branches have no common ancestor, and no merge base revision was specified.
+ /etc/init.d/bcfg2-server restart
Stopping Configuration Management Server: * bcfg2-server
Starting Configuration Management Server: * bcfg2-server
+ netstat -atn
+ grep :6789
+ [ -z ]
+ sleep 5
+ netstat -atn
+ grep :6789
+ [ -z tcp 0 0 0.0.0.0:6789 0.0.0.0:* LISTEN ]
+ echo 1 minute: BCFG2 is loading packages and config files
1 minute: BCFG2 is loading packages and config files
+ sleep 60
+ bcfg2-admin xcmd Packages.Refresh
+ bcfg2 -vqe
Server failure: The read operation timed out
Failed to download probes from bcfg2
The read operation timed out
+ /etc/init.d/varnish stop
 * Stopping HTTP accelerator
   ...fail!
+ ls /etc/mercury/boot.d/S050ec2.sh /etc/mercury/boot.d/S100main.sh
+ bash /etc/mercury/boot.d/S050ec2.sh
mv: cannot move `/var/log/mysql' to `/mnt/mysql/log': No such file or directory
ln: creating symbolic link `/var/log/mysql/log': File exists
mv: cannot move `/var/lib/mysql' to `/mnt/mysql/lib': No such file or directory
ln: creating symbolic link `/var/lib/mysql/lib': File exists
mv: cannot move `/var/lib/varnish' to `/mnt/varnish/lib': No such file or directory
ln: creating symbolic link `/var/lib/varnish/lib': File exists
chown: cannot access `/mnt/varnish/lib/pressflow/': No such file or directory
chmod: cannot access `/etc/init.d/ec2-ssh-host-key-gen': No such file or directory
+ bash /etc/mercury/boot.d/S100main.sh
 * Stopping Postfix Mail Transport Agent postfix
   ...done.
 * Starting Postfix Mail Transport Agent postfix
   ...done.
+ /etc/init.d/mysql start
Rather than invoking init scripts through /etc/init.d, use the service(8)
utility, e.g. service mysql start

Since the script you are attempting to invoke has been converted to an
Upstart job, you may also use the start(8) utility, e.g. start mysql
+ /etc/init.d/varnish start
 * Starting HTTP accelerator
   ...fail!
Error: (-sfile) "/var/lib/varnish/pressflow/varnish_storage.bin" does not exist and could not be created
+ echo create database pantheon;
+ mysql -u root
ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)
+ date
+ echo Mon Jul 19 21:19:01 UTC 2010
+ echo

##############################
# Mercury Setup Complete! #
##############################

##############################
# Mercury Setup Complete! #
##############################

+ wall
+ echo
DEAR SYSADMIN: MERCURY IS READY FOR YOU NOW

Do not forget the README.txt, CHANGELOG.txt and docs!

Revision history for this message
Greg Coit (gregcoit) wrote :

I'm unable to reproduce this error. On both the 32 and 64 bit version of the Mercury 1.1 Beta 2 (ami-294aa340 and ami-554aa33c), the init script runs fine except for this line (which is related to http://groups.drupal.org/node/80324):

Merging from remembered parent location http://bazaar.launchpad.net/~pantheon-developers/pantheon/1.1/
bzr: ERROR: Branches have no common ancestor, and no merge base revision was specified.

The solution to this error is:

/etc/init.d/bcfg2-server stop
sudo rm -rf /var/lib/bcfg2/
sudo bzr branch lp:pantheon/1.1 /var/lib/bcfg2
echo -e "<Clients version=\"3.0\">\n</Clients>" | sudo tee -a /var/lib/bcfg2/Metadata/clients.xml
sudo /etc/init.d/bcfg2-server start; tail -f /var/log/syslog
sudo bcfg2 -vqed

As for errors reported by init.sh on your server, i think they are caused by the second instance of init.sh being started while the original (system started) init.sh was still running.

Changing to invalid.

Changed in pantheon:
status: New → Invalid
Revision history for this message
Pooya Khalili (pooya-khalili) wrote :

Hi.. I am getting a similar error, I tried the steps you mentioned and and tried mercury.init > Build Now in hudson, but I still get these errors. specificaly towards the end, this is what I get:

Failed to manipulate service dmesg
Phase: final
Correct entries: 490
Incorrect entries: 2
 Service:network-interface Service:ureadahead
Total managed entries: 492
Unmanaged entries: 257

Mercury Update Complete!
Starting Pressflow Update Script
Using saved parent location: http://bazaar.launchpad.net/~pressflow/pressflow/6/
No revisions to pull.
Pressflow Update Complete!
Starting AWS Specific Configuration (non ebs-boot instances only)
mysql stop/waiting
mv: cannot overwrite directory `/mnt/mysql/lib/mysql' with non-directory
mysql start/running, process 7572
 * Stopping HTTP accelerator
   ...done.
 * Starting HTTP accelerator
   ...done.
chmod: cannot access `/etc/init.d/ec2-ssh-host-key-gen': No such file or directory
Finished: FAILURE

and then when I check to see if mercury is running it's not. any help would be appreciated.. thanks

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.