Nagios Core 4.3.4 - NCPA passive Checks - SE Linux

[charangandra]

Hi,

I've installed Nagios core on Centos 7.3 and using NCPA for passive checks. NCPA is able to send the check results back to nrdp as long as I disabled SE Linux. But Once I enable the SE Linux it is unable to send the details and giving the below error.

SE Linux Enabled: Error as below

2018-08-31 14:58:14,946 108325 INFO Message from NRDP server: BAD CHECK RESULTS DIR
2018-08-31 14:58:14,946 108325 INFO Meta output from NRDP server: Nonexistent

SE Linux Disabled: usual log below

2018-08-31 14:50:07,567 108325 INFO Running check: disk/logical/| --warning 80 --critical 90 --units Gi
2018-08-31 14:50:07,622 108325 INFO Running check: cpu/percent --warning 80 --critical 90 --aggregate avg
2018-08-31 14:50:08,184 108325 INFO Running check: memory/swap --warning 80 --critical 90 --units Gi
2018-08-31 14:50:08,246 108325 INFO Running check: memory/virtual --warning 80 --critical 90 --units Gi
2018-08-31 14:50:08,309 108325 INFO Running check: processes --warning 300 --critical 400
2018-08-31 14:50:08,544 108325 INFO Message from NRDP server: OK
2018-08-31 14:50:08,545 108325 INFO Meta output from NRDP server: 5 checks processed.

Do we need to definitely disabled the SE Linux? It is project requirement to have SE Linux enabled, how can I resolve the issue?

Thanks,
Charan

[scottwilkerson]

Do we need to definitely disabled the SE Linux? It is project requirement to have SE Linux enabled, how can I resolve the issue?

Yes, you either need to disable selinux, or write a policy that will allow apache to have access to the check results directory

[charangandra]

Thanks for your quick reply. Is there any Link to follow to write a policy that will allow apache to have access to the check results directory?

Thanks,

[scottwilkerson]

We don't have any publications on the topic, but a quick Google search found the following
http://iris77.net/?p=585