Apply Settings

[wiproltdwiv]

Hi,

we had around 15000+ hosts and services under monitoring.

our problem is while adding new devices , apply settings
1. it takes too much time to complete the configuration even for adding single host.
2. most of the time we are getting configuration error like ' import data failed' and there is no configuration snap shot for the same
3. post editing any hosts in admin console.. the apply settings took min 30 min to complete the configuration.
4.during this apply settings , is all the host and services reconnect or re establish (or) only specific host which we added newly get updated. ?

pls suggest for better montoring for all the above 4 points

[abrist]

1. it takes too much time to complete the configuration even for adding single host.
2. most of the time we are getting configuration error like ' import data failed' and there is no configuration snap shot for the same

How may checks every five minutes are you running?

What are your php.ini limits set to:

Code: Select all

grep "max_execution_time =\|memory_limit =\|max_input_time =" /etc/php.ini

3. post editing any hosts in admin console.. the apply settings took min 30 min to complete the configuration.

What are the specs on the Nagios XI server?
When you apply configuration, run "top" from a terminal and check the "x.x%wa" value at the top - this is the i/o wait value and many times is the culprit.

4.during this apply settings , is all the host and services reconnect or re establish (or) only specific host which we added newly get updated. ?

*ALL* config files are written out when "apply configuration" is run.

[wiproltdwiv]

for point 1 & 2

grep "max_execution_time =\|memory_limit =\|max_input_time =" /etc/php.ini
#max_execution_time = 220
max_execution_time = 520
max_input_time = 120
memory_limit = 2048M

we are monitoring around 5000+ routers and links in the tool

[slansing]

What are the specs on the Nagios XI server?
When you apply configuration, run "top" from a terminal and check the "x.x%wa" value at the top - this is the i/o wait value and many times is the culprit.

How may checks every five minutes are you running?

Please answer the above. The Nagios XI server will check all of the configuration files to verify the integrity of them, and verify them against the last known configuration. This also assists in connecting one object to another if a change such as this occurs. In addition, how many services do you have? 10,000? Going off your original number and subtracting the one above.

[wiproltdwiv]

hi,

yes we are using 10,000 checks for every five minutes and more over in future it will increase.

for your info: we purchased unlimited host and service license.

please suggest , Tool is struggling to add hosts and services....

Attached top command output for your reference

[abrist]

I have a number of questions after looking at the screens and the rest of the thread:
1. The check_oracle_health plugins are using quite a bit of cpu - what type of query are they doing?
2. Apache in the second image is using 99% CPU - do you have a large number of scheduled reports running?
3. How many users are using the web ui frontend at one time?
4. Are you using any high performance tweaks? (rrdcache, memdisk, sql offload)
5. What are the system specs or provisioning for the nagios server?

[wiproltdwiv]

below are update on as per given points

1) we have 33 oracle db which are using check_oracle_health plugin and for one db we are monitoring approx. 35 services.

2) No. currently we did not scheduled any report.

3) 3 users accessing console with different locations at a time.

4) Yes, we are using rrdcache as well we have upload sql db on remote server. We have two servers with same configuration as below, one for tool and rest for sql db.

5) We are using RHEL 6.3 (64 bit). CPU and Memory details are attached.

[abrist]

Alright. Specs look good. DO you see the httpd service spike frequently (resembling your screenshot)?

[wiproltdwiv]

No

[sreinhardt]

Going back to your first post, do you still have items that are attempting to be imported? They would show in /usr/local/nagios/etc/import/