Category Archives: SysAdmin

Postfix in Internet, Relay in LAN

Assume the following scenario:

Your Postfix MTA is located in the Internet (having a public IP address) and you have configured an external DNS server there to be used for DNS requests.  But you are going to send mails using an internal mail server as mail relay. In this case the private IP address of the internal mail relay cannot be re-solved by the external DNS server. In this situation Postfix cannot use the mail relay and shows an error message like this below in the log file:

Aug 26 12:49:12 myclient postfix/error[28425]: B35AF34B: to=<me@company.local>, relay=none, delay=0.14, delays=0.09/0/0/0.04, dsn=4.3.5, status=deferred (delivery temporarily suspended: Host or domain name not found. Name service error for type=AAAA: Host not found)

To re-solve the issue you have to add the following statement to

disable_dns_lookups = yes

and re-load postfix:

service postfix reload

Use the appropriate command to enforce delivery of queued mails.


Some useful postfix commands

To re-try delivering of queued mails:

postqueue -f

To remove all deferred mails:

postsuper -d ALL deferred

To remove all mails

postsuper -d ALL

Insufficient system resources (in reply to MAIL FROM command)

If you see an error like this while trying to send an email via your Exchange server

E32F1291C 5322 Tue Aug 26 09:15:08 MAILER-DAEMON
 (host[] said: 452 4.3.1 Insufficient system resources (in reply to MAIL FROM command))

then the system partition of the Exchange server has become low on free space. I have been watching this, if the free space on the drive C: of my server becomes less then 6% of the whole partition size. These are still more than 3 GByte!

You don’t need to do anything on the Exchange server but to free some additional space.


OMD: If PNP4Nagios XML file has changed

Assume the following situation:

You changed something on a service and the performance data format has changed as well. PNP4Nagios graphs that belong to the changed service will not be changed automatically. You have to remove the old XML file that belongs to the changed service and then re-start your OMD site completely like this:

OMD[mysite]:~$ rm var/pnp4nagios/perfdata/webserver35/Apache_127.0.0.1_80_Status.xml
OMD[mysite]:~$ omd restart

In a few minutes the XML file should be re-generated and graphs belonging to new performance data counters should appear. If NOT, I’m afraid that you have to remove all RRD-files that belong to the service (you can recognize it on the file name) and to hope that they will be re-created properly. In this case all the server statistic will be lost (you have been warned!).

Copying a site in OMD

OMD provides some very useful commands to manage sites. Execute “omd help” as root to see all possible commands.

Continue reading

Exchange 2010 Mailbox Database Info Check for Check_MK updated

I’ve updated the check and built a new package. The package can be downloaded from Check_MK exchange site here. The new version has the number 1.4.

There are two changes regarding the agent’s side: the plugin call logic and the caching of results.

Continue reading

Extend a Logical Volume to 100% of Volume Group

To extend a logical volume (LV) named ´logs` to the whole space of the volume group (VG) named ´data` execute the following command:

# lvextend -l 100%VG /dev/data/logs

Fortigate checks for Check_MK updated

Today I’ve uploaded a new version of checks for Fortigate devices. The new version is 2.2 and can be downloaded from the exchange site of the Check_MK project. These are new features:

  • added a check to monitor the synchronization status of Fortigate high availability clusters (HA cluster)
  • added a check to monitor the number of current VPN SSL connections
  • added perf-o-meters for some checks
  • minor changes

The checks have worked under CMK 1.2.4p2 and 1.2.4.p3, but might work with other versions of Check_MK. The Fortigate devices the checks work successfully with are: Fortigate 60C/60D/90D/300C