Page 1 of 2

The LUN mapped for Nagios is utilizing huge IO

Posted: Thu Mar 05, 2015 8:32 am
by wiproltdwiv
The LUN mapped for Nagios is utilizing huge IO

LUN Total OP/s Application running
/vol/EMSAPM2/emsapmdata 745 Nagios

recently added around 1600 parameters in Nagios server, is this the reason for IO utilization. Please help & suggest. :oops:

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Thu Mar 05, 2015 1:28 pm
by ssax
It depends on what type of parameters you are using that would determine the amount of IO used.

Have you added a RAMDisk? You can also read more about the the Key Files That Affect Disk I\O in the following document:

http://assets.nagios.com/downloads/nagi ... giosXI.pdf

For good measure I'll link you to the maximizing performance doc:

http://assets.nagios.com/downloads/nagi ... ios-XI.pdf

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Thu Mar 12, 2015 12:27 am
by wiproltdwiv
What are the parameters that can take more efficency or more IO,

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Thu Mar 12, 2015 12:44 am
by wiproltdwiv
I mean how should I check which parameter is consuming more IO, so that we can find the culprit & remove... :oops:

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Thu Mar 12, 2015 12:56 am
by Box293
iotop should help with this

Code: Select all

yum -y install iotop
then run

Code: Select all

iotop

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Fri Mar 13, 2015 2:10 am
by wiproltdwiv
Is the files creating in /usr/local/nagios/var/spool/xipde are neccessary for any work..because it is around 90G in our nagios server..pls suggest can we remove it.. :oops:

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Fri Mar 13, 2015 5:26 am
by wiproltdwiv
Can we delete these files...Will it affect ???

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Fri Mar 13, 2015 9:19 am
by ssax
That directory contains all the performance data, if you delete it you will lose your past performance data.

90G is a lot of space, should really be around 1-2G

If you don't care about the performance data you can delete the files but you should still modify the file below and change your settings.

Change /usr/local/nagios/etc/pnp/npcd.cfg
From

Code: Select all

load_threshold = 10.0
To: *** A general rule for load_threshold would be the number of cores your server has times 10.

Code: Select all

load_threshold = 20.0
Then run:

Code: Select all

service npcd restart
If that doesn't clean up that directory you may have to delete the files and then restart the npcd service.

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Wed Mar 18, 2015 7:22 am
by wiproltdwiv
/usr/local/nagios/var/spool/xidpe/ This file is 15 G size, I am not even able to tar or zip it.showing errors like :
tar: /usr/local/nagios/var/spool/xidpe/ziLwfhAq: File removed before we read it
tar: /usr/local/nagios/var/spool/xidpe: file changed as we read it

Requesting you to please help us with any suggestion :oops:

Re: The LUN mapped for Nagios is utilizing huge IO

Posted: Wed Mar 18, 2015 11:07 am
by abrist
You will most likely have to remove the files in the xidpe folder as the spool has gotten too large to be stat()ed/parsed. How many files are in there:

Code: Select all

ls /usr/local/nagios/var/spool/xidpe/ | wc -l
You can remove the files with the following command, but you should be aware that you will have gaps in your graphs representing the time associated with each of these performance data files in xidpe:

Code: Select all

cd /usr/local/nagios/var/spool/xidpe/
find . -type f -delete