Setting up pulp for update managment`

[steevod]

Hi All

I'm looking to setup pulp for patch/update managment. I've managed to install it OK, along with the many dependencies. I'm using CentOS 5.5 on a VMWare virtual machine.

When I try to start it for the first time, I get the following...

# service pulp-server start
Pulp database not initialized, please run: /etc/init.d/pulp-server init

...so I start it with the init parameter...

# service pulp-server init
Starting mongod: [ OK ]
Traceback (most recent call last):
File "/usr/bin/pulp-migrate", line 19, in ?
from pulp.server.db.migrate import script
File "/usr/lib/python2.4/site-packages/pulp/server/db/migrate/script.py", line 27, in ?
connection.initialize()
File "/usr/lib/python2.4/site-packages/pulp/server/db/connection.py", line 42, in initialize
_connection = pymongo.Connection(seeds)
File "/usr/lib64/python2.4/site-packages/pymongo/connection.py", line 184, in __init__
self.__find_master()
File "/usr/lib64/python2.4/site-packages/pymongo/connection.py", line 456, in __find_master
raise AutoReconnect("could not find master")
pymongo.errors.AutoReconnect: could not find master

...I suspect this is because pulp uses mongodb, and mongodb doesn't look to be starting properly. When I start it, I get the following...

# service mongod start
Starting mongod: [ OK ]

...which looks OK, but the log file reports...

Tue Jan 25 13:01:11 MongoDB starting : pid=13069 port=27017 dbpath=/var/lib/mongodb 64-bit
Tue Jan 25 13:01:11 db version v1.6.4, pdfile version 4.5
Tue Jan 25 13:01:11 git version: nogitversion
Tue Jan 25 13:01:11 sys info: Linux x86-11.phx2.fedoraproject.org 2.6.32-71.el6.x86_64 #1 SMP Wed Sep 1 01:33:01 EDT 2010 x86_64 BOOST_LIB_VERSION=1_33_1
Tue Jan 25 13:01:11 [initandlisten] Accessing: local for the first time
Tue Jan 25 13:01:11 [initandlisten] query local.system.namespaces reslen:36 nreturned:0 1ms
Tue Jan 25 13:01:11 Got signal: 11 (Segmentation fault).

Tue Jan 25 13:01:11 Backtrace:
0x7b5b14 0x36736302d0
/usr/bin/mongod(_ZN5mongo10abruptQuitEi+0x384) [0x7b5b14]
/lib64/libc.so.6 [0x36736302d0]

Tue Jan 25 13:01:11 dbexit:

Tue Jan 25 13:01:11 [initandlisten] shutdown: going to close listening sockets...
Tue Jan 25 13:01:11 [initandlisten] shutdown: going to flush oplog...
Tue Jan 25 13:01:11 [initandlisten] shutdown: going to close sockets...
Tue Jan 25 13:01:11 [initandlisten] shutdown: waiting for fs preallocator...
Tue Jan 25 13:01:11 [initandlisten] shutdown: closing all files...
Tue Jan 25 13:01:11 closeAllFiles() finished

Tue Jan 25 13:01:11 [initandlisten] shutdown: removing fs lock...
Tue Jan 25 13:01:11 dbexit: really exiting now

...so it looks like it tries to start, and then gets a segmentation fault and shuts down.

Has anyone had any experience in setting this up? Any ideas as to the cause of the problem?

[kholmes]

I'm currently working on setting this up to test in a newish cluster. Where did you get your qpid-cpp-server* dependencies? Perhaps if I get past this I can see if I encounter the same problem

[steevod]

No problem, trawling through all the dependencies was a real PITA. My post count isn't high enough to post them here, so I've pm'd the URL's to you.

[kholmes]

Thanks for the links, they worked. I was getting errors about a missing directory /var/lib/mongodb. I created that and set up the permissions properly so 'mongodb' was the owner. You could check the permissions, they might be incorrect?