Zimbra offers Open Source email server software and shared calendar for Linux and the Mac
Go Back   Zimbra :: Forums > Zimbra Collaboration Suite > Administrators

Welcome to the Zimbra :: Forums!
Welcome, if you would like to post a comment please register. We also encourage you to explore all things Zimbra with our team and members of the community.

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
  #1 (permalink)  
Old 07-17-2011, 04:20 AM
xda xda is offline
New Member
 
Posts: 3
Default Zimbra configd not starting

Code:
Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Starting zmconfigd...Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Starting zmconfigd...failed.
Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Starting zmconfigd...Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
failed.
Starting zmconfigd...failed.
Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Starting zmconfigd...Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
failed.
Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Starting zmconfigd...failed.
Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
Starting zmconfigd...Traceback (most recent call last):
  File "/opt/zimbra/libexec/zmconfigd", line 149, in <module>
    Log.logMsg(1, "%s started on %s with loglevel=%d pid=%d" % (myConfig.progname, myConfig.hostname, myConfig.loglevel, os.getpid()))
AttributeError: 'module' object has no attribute 'getpid'
I couldn't get it working, so I tried using the solutions used by others. I am running zimbra on CentOS 5.5 64-bit.

I deleted the .pid file, executed the zmfixperms. Still no luck, all other services are running including SNB after I installed the patch. The only thing not running is zmconfigd. Even sending and receiving e-mails is working.

Anyway how can I rebuild the .pid, get it back , because it looks like that caused a major problem.

Regards,
XDA
Reply With Quote
  #2 (permalink)  
Old 07-17-2011, 04:44 AM
Zimbra Consultant & Moderator
 
Posts: 20,314
Default

Quote:
Originally Posted by xda View Post
[code]I couldn't get it working, so I tried using the solutions used by others. I am running zimbra on CentOS 5.5 64-bit.
Is this a new install or an upgrade? If you have valid DNS A & MX records and a correct /etc/hosts file there really should be no problems installing on RHEL (CentOS) systems - do you have those?
__________________
Regards


Bill
Reply With Quote
  #3 (permalink)  
Old 07-17-2011, 05:55 AM
xda xda is offline
New Member
 
Posts: 3
Default

Quote:
Originally Posted by phoenix View Post
Is this a new install or an upgrade? If you have valid DNS A & MX records and a correct /etc/hosts file there really should be no problems installing on RHEL (CentOS) systems - do you have those?
This is a new install! I have valid DNS A & MX records. My /etc/hosts file contains all necescary information. Btw. why would I be able to receive and send e-mail if DNS was not setup properly...

Anyway the onlything that I can think off, that is giving the problems is the CSF firewall, but I already opened up all ports that are required by zimbra.

And isn't the .pid file supposed to be generated automaticly after each restart?

Regards,
xda
Reply With Quote
  #4 (permalink)  
Old 07-17-2011, 06:26 AM
Zimbra Consultant & Moderator
 
Posts: 20,314
Default

Is this actually running on on bare metal or in a VM and if so, which VM and how much RAM does this server have? Have you also applied the required patch to this version of Zimbra (you should, it's on the download page)?
__________________
Regards


Bill

Last edited by phoenix; 07-17-2011 at 06:38 AM..
Reply With Quote
  #5 (permalink)  
Old 07-17-2011, 06:42 AM
xda xda is offline
New Member
 
Posts: 3
Default

Quote:
Originally Posted by phoenix View Post
Is this actually running on on bare metal or in a VM and if so, which VM and how much RAM does this server have? Have you also applied the required patch to this version of Zimbra (you should, it's on the download page)?
Yes, I applied the patch. It is a private cloud network. It runs Xen instances, we allocated 24 Gigabyte of RAM and 16 cores to the current instance. The only things installed before installation were Nginx, PHP-FPM, Python 2.7, Django, PHP 5.3, MySQL, PostgreSQL, CSF Firewall, PHP-XCache, Memcached, NSD, ClamAV (but turned off, using DynDNS now).
Reply With Quote
  #6 (permalink)  
Old 07-21-2011, 08:53 AM
Junior Member
 
Posts: 6
Default

For starters, that's not 'only things'; that's a complete software stack, almost none of which should be present in a clean Zimbra install, and a lot of which will actually interfere with Zimbra.

Having said that, I'm also having this issue on a CentOS 5.6 machine.

Check:
Issue 1518: os.getpid() is missing - Jython tracker

The simple workaround for now is to stop Zimbra completely and replace /opt/zimbra/lib/jars/jython-2.5.2.jar with the jython-2.5.1.jar that came with the original 7.1.1 setup (you can extract it from zimbra-core-<whatever>.rpm using rpm2cpio), that works for me.

Zimbra devs: Is this a known issue yet? I can't find it in the Bugzilla yet, but it's a clear and very reproducable bug.
Reply With Quote
  #7 (permalink)  
Old 07-21-2011, 09:16 AM
Junior Member
 
Posts: 6
Default

Code:
[root@zimbra-test jars]# cp /home/admin/ZCS/jython-2.5.1.jar /opt/zimbra/lib/jars/ ; rm -f /opt/zimbra/lib/jars/jython-2.5.2.jar
##########
[zimbra@zimbra-test.example.com ~]$ zmpython
Jython 2.5.1 (Release_2_5_1:6813, Sep 26 2009, 13:47:54) 
[Java HotSpot(TM) 64-Bit Server VM (Sun Microsystems Inc.)] on java1.6.0_25
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> os.getpid()
0
>>> exit()


Code:
[root@zimbra-test jars]# cp /home/admin/ZCS/jython-2.5.2.jar /opt/zimbra/lib/jars/ ; rm -f /opt/zimbra/lib/jars/jython-2.5.1.jar
##########
[zimbra@zimbra-test.example.com ~]$ zmpython
Jython 2.5.2 (Release_2_5_2:7206, Mar 2 2011, 23:12:06) 
[Java HotSpot(TM) 64-Bit Server VM (Sun Microsystems Inc.)] on java1.6.0_25
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> os.getpid()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'getpid'
>>>
Reply With Quote
  #8 (permalink)  
Old 07-21-2011, 12:11 PM
Junior Member
 
Posts: 6
Default

Well, turns out the devs do know about this issue

SourceForge jython-users

So there's a few possibilities regarding this bug

* zmpython is supposed to redirect everything to within /opt/zimbra, yet for some reason, 2.5.2 (and apparently not 2.5.1) uses /tmp to write stuff in, not /opt/zimbra/data/tmp
* consequently, if something prevents Jython from executing within /tmp, this will occur
** SELinux (but if you've RTFM'ed, you've got that disabled)
** mounting /tmp noexec will do the same trick, which turns out to be my problem.

Again, however, 2.5.1 doesn't have this problem

Last edited by The Loeki; 07-21-2011 at 12:24 PM..
Reply With Quote
  #9 (permalink)  
Old 07-21-2011, 12:26 PM
Junior Member
 
Posts: 6
Default

Of course the real issue is why Jython 2.5.2 will apparently still try to write to /tmp, while 2.5.1 doesn't.
Reply With Quote
  #10 (permalink)  
Old 07-22-2011, 09:19 AM
Junior Member
 
Posts: 6
Default

And figured that out too. Long story short, 2.5.2 uses JFFI/JAFFL instead of JNA, which is why it's writing to /tmp.

Specifying -Djava.io.tmpdir=/opt/zimbra/data/tmp in zmpython fixes the problem permanently; an even better one would probably be to add it to zmjava's default options.
Reply With Quote
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes


Similar Threads

Why Join?

Registering let's you ask questions, makes it easier to search, displays any files attached to posts, and notifies you about replies.

blog.zimbra.com




 

SEO by vBSEO ©2011, Crawlability, Inc.