puppet logs are missing in /var/log/remote (and UI as well)

Bug #1316093 reported by Tatyanka
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Fuel for OpenStack
Confirmed
High
Bogdan Dobrelya

Bug Description

{"build_id": "2014-04-30_10-31-20", "mirantis": "yes", "build_number": "170", "ostf_sha": "134765fcb5a07dce0cd1bb399b2290c988c3c63b", "nailgun_sha": "fb676be97794d4b63b8d179e2a8518ad70448f48", "production": "docker", "api": "1.0", "fuelmain_sha": "44954d8ffd74d27219434ec5b676874504b6bf76", "astute_sha": "3cffebde1e5452f5dbf8f744c6525fc36c7afbf3", "release": "5.0", "fuellib_sha": "fd547e5008957183200608124c677c096bf099f9"}

Deploy HA with Centos
When Openstack starts to deploy on node - go to UI and try to find puppet logs

Actual result:
There is no logs from puppet (see screen) But puppet apply is running
sysconfig for puppet looks like:
# The puppetmaster server
#PUPPET_SERVER=puppet

# If you wish to specify the port to connect to do so here
#PUPPET_PORT=8140

# Where to log to. Specify syslog to send log messages to the system log.
#PUPPET_LOG=/var/log/puppet/puppet.log

# You may specify other parameters to the puppet client here
#PUPPET_EXTRA_OPTS=--waitforcert=500

rsyslog condifg looks like:
# file is managed by puppet

### collect puppet logs of all levels in /var/log/puppet/puppet.log
if ($programname == 'puppet-apply' \
 or $programname == 'puppet-user' \
 or $programname == 'puppet-error') \
### use 6 for INFO+ msg levels, 5 for WARN+ etc.
### and $syslogseverity <= '6' \
then -/var/log/puppet/puppet.log
### stop further processing for the matched entries
& ~

root@node-7 ~]# ps afuxww | grep puppet
root 30834 0.0 0.0 103244 852 pts/0 S+ 11:21 0:00 \_ grep puppet
root 7975 9.2 28.2 548852 434072 ? Ss 11:10 1:00 /usr/bin/ruby /usr/bin/puppet apply /etc/puppet/manifests/site.pp --logdest syslog --trace

Revision history for this message
Tatyanka (tatyana-leontovich) wrote :
Changed in fuel:
milestone: none → 5.0
Changed in fuel:
status: New → Incomplete
importance: Undecided → Low
assignee: nobody → Bogdan Dobrelya (bogdando)
Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

Please provide additional info
1) Logs snapshot (it is important to see how the events flow in the time)
2) Can you see the puppet-apply.log in /var/log/remote/node-* while it is absent in UI?
3) Then (and if) puppet logs would appear in UI?

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

4) What is the /var/log/puppet content for the node-7 (the one which has a puppet running) ?
5) Does ctrl+f5 resolve the UI issue?

Revision history for this message
Tatyanka (tatyana-leontovich) wrote :

2) Can you see the puppet-apply.log in /var/log/remote/node-* while it is absent in UI? -
on controllers:
[root@node-7 ~]# ls -all /var/log/puppet/
total 260
drwxr-x---. 2 puppet puppet 4096 May 5 11:11 .
drwxr-xr-x. 20 root root 4096 May 5 11:20 ..
-rw------- 1 root adm 251068 May 5 11:31 puppet.log

On other nodes (compute/ceph/cinder puppet logs show on UI without problem)

 Then (and if) puppet logs would appear in UI? - no
5) - Does ctrl+f5 resolve the UI issue? - no
Al other details I hope you can find in attached snapshot

Revision history for this message
Tatyanka (tatyana-leontovich) wrote :
Changed in fuel:
status: Incomplete → Confirmed
importance: Low → Medium
Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

Cannot reproduce it on #180 with Centos HA 1 controller + 1 compute

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

The issue could be related to the recently fixed docker container issues

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

I mean this one #1315865

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

If it is possible to access the env with the issue reproduced on it, please provide an info about
1) status of docker-rsyslog in supervisorctl
2) status of ss -untap | grep 514
while controller nodes are being deployed

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

I found that I have both (1), (2) in the given diagnostic snapshot and it shows:
1) there is no docker started for rsyslog, see ./localhost/commands/ps.txt
2) there is nothing on 514 port to listen for logs, see ./localhost/commands/netstat.txt
So, the source of this issue is an incorrect state of docker container for rsyslog.
The correct state should be like this http://pastebin.com/EtSJASBG

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

Once I cannot reproduce it on #180, I set this one as a dup of 1315865.

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

Apparently reproduced, see details in logs. This time tcpdump captuted pupet logs and local logs for nodes included. Note: both controller and compute nodes are missing pupet-apply logs in /var/log/remote, hence in UI as well.

Revision history for this message
Bogdan Dobrelya (bogdando) wrote :
  • logs Edit (6.2 MiB, application/x-tar)
Changed in fuel:
importance: Medium → High
summary: - puppet logs missed in UI
+ puppet logs missed in /var/log/remote (and UI as well)
summary: - puppet logs missed in /var/log/remote (and UI as well)
+ puppet logs are missing in /var/log/remote (and UI as well)
Revision history for this message
Bogdan Dobrelya (bogdando) wrote :

Yes, I was able to reproduce it only after some env deletions and resets. So I set it as a dup

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.