Nagios XI Slow Web interface.

[jonatgil]

Hello,
We are currently in process of implementing a new instance of our Nagios Xi from scratch.
And I find the web interface pretty slow and I was wondering what I could do to make that better before telling everyone to connect on it.
1500 hosts. 160 services. 6000 total checks. Zero notification. Nagios XI 5.6.0

Will provide profile or other info if needed.
4 Core 6 GB Ram.

Not sure why Buffers and Cached memory are high?

Here is was I did so far, by looking at older similar post.
/etc/php.ini -> max_input_vars = 10000
/etc/my.cnf -> max_connections=500


Memory
Total 5,796 MB
Used 979 MB
Free 336 MB
Shared 307 MB
Buffers 4,480 MB
Cached 4,174 MB

Load
1-min 0.97
5-min 1.31
15-min 1.45
CPU Stats
User 31.67%
Nice 0.00%
System 9.42%
I/O Wait 0.15%
Steal 0.00%
Idle 58.76%

Receiving the same output when loading page from admin account.

[root@XXXXXXXX ~]# ps aux | awk '{print $6 " " $11}' | sort -nr | head -20
371864 /usr/libexec/mysqld
68912 /usr/sbin/rsyslogd
45712 /usr/sbin/httpd
33240 /usr/bin/python
33156 /usr/sbin/httpd
32964 /usr/sbin/httpd
32876 /usr/bin/php
32788 /usr/bin/php
32504 /usr/sbin/httpd
29680 /usr/sbin/httpd
28612 /usr/sbin/httpd
28544 /usr/sbin/httpd
28432 /usr/sbin/httpd
28396 /usr/sbin/httpd
28128 /usr/sbin/httpd
28116 /usr/sbin/httpd
25876 /usr/bin/php
25824 /usr/bin/php
25636 /usr/bin/php
24920 /usr/bin/php

Let me know.
Thank you

[jonatgil]

Would a ram disk help for that matter?

More details...

Metric
Value

Host Check Latency
Min 0.00 sec
Max 1.06 sec
Avg 0.01 sec

Host Check Execution Time
Min 0.00 sec
Max 10.01 sec
Avg 0.46 sec

Service Check Latency
Min 0.00 sec
Max 1.04 sec
Avg 0.01 sec
Service Check Execution Time
Min 0.00 sec
Max 60.01 sec
Avg 1.40 sec

[jonatgil]

Just change the Reaper Settings from the Maximizing-Performance-In-Nagios-XI.pdf and it seems to be better now.

Modified reaper setting for a high check volume:
check_result_reaper_frequency=3
max_check_result_reaper_time=10

[benjaminsmith]

Hi @jonatgil,

Just change the Reaper Settings from the Maximizing-Performance-In-Nagios-XI.pdf and it seems to be better now.

It looks like you were able to make improvements. You might also consider enabling Unified Pages. Go to Admin > System Config > Performance Settings > Pages. It's explained in the Maximizing Performance In Nagios XI guide.

[jonatgil]

Would the ram disk would be only useful in the case of high latency?

''he biggest hardware limitation to a Nagios system is disk I/O ''
'' This can cause check latencies to soar even though the CPU usage appears within a safe range''
Or it could help the overall speed of the site?
Thank you.


If I understand correctly the ''Unified Pages'' will only affect page/section of page that a lot of people are requesting?
it should not have impact outside of that?
Thank you.

[benjaminsmith]

Hi @jonatgil,

Yes. Nagios Core generates a significant amount of disk activity to update status files, performance data and logs and adding a RAM disk will help speed up this process. However, I would first recommend increasing the RAM and CPU cores on your server ( if possible ). Our recommendations for systems with more than 500 hosts and 2500 service checks is at least 4 CPU cores and 8 GB of RAM.

Nagios XI Hardware Requirements

That's correct the Unified pages will help speed up certain pages in the interface by decreasing the number of AJAX calls.

[jonatgil]

The actual environment was not slow when we transfer it to the prod network.
I think that since 1/3 of the hosts are inside a special network/DMZ there a lot of timeout/waiting for response.

[scottwilkerson]

The actual environment was not slow when we transfer it to the prod network.
I think that since 1/3 of the hosts are inside a special network/DMZ there a lot of timeout/waiting for response.

That makes sense, and explains why you were hitting the max 60secs on some services based on the Max Service Check Execution Time number