Loading page Monitoring-Services extremely slow

[ssoliveira]

Hello good afternoon.

We have an old problem in our environment of Nagios XI, that every day becomes slower.

When a user accesses the "Monitoring >> Services" page, this is extremely slow.

We realize that slowness only occurs when there are no filters, or searches.

If you search for a service, or select, the load is normal.

It seems; the page reads all the records in the database; to later return the data; and perform paging (it seems that there is no paging logic on the database side).

I would like to see a logic implemented that addresses this problem.

I believe that: If the page is modified (adding a logic that checks for a filter or search, and if not, return the first services), solve this problem.

We are using version "5.4.12" that has been studied to be updated. But we did not find any notes regarding this issue in release-notes.

[benjaminsmith]

Hello @ssoliveira,

Thanks for uploading the images - very helpful. We've done some testing, and I have a couple of quick questions that would help us out.

1. How many hosts and services do you have?

2. How many seconds does it take to load the next page? Also, do you know notice any difference in page loads with you change the results per page setting from 15 to a higher value (e.g. 100, 250)

Please let me know.

In the meantime, increasing the memory limit in /etc/php.ini may help.
https://support.nagios.com/kb/article/n ... e-611.html

[ssoliveira]

We have 1,700 hosts, and 59,100 services. The workload is distributed over 5 gearmans. Only network monitors, such as mrtg, are in the core because they can not be moved because of the way monitoring occurs. Mysql is on a dedicated, seperate server.

The server works with low CPU and disk consumption. Our last optimization was to add LUN disks via SAN in Nagios, from an all-flash Storage vmax to not have any IO bottleneck. Currently the infrastructure has 20,000 IOPS available, but consumes very little.

Without setting filters, the page takes from 40 to 60 seconds. As the page times out, some people end up trying to open the "right-click, open another page" page several times (while the first request has not yet ended), which eventually has generated a need to restart apache.

The number of results per page does not change the behavior. The times of editing the services, and loading when they have filters are practically instantaneous.

The loading time for the next pages is the same as the initial load. If I open the page, it takes about 50 seconds; if I click on number 2, to go to the next page, it will take another 50 seconds. etc.

The memory limit is currently at 2GB, but this does not change the behavior.

Code: Select all

; Maximum amount of memory to script may consume (128MB)
; http://www.php.net/manual/en/ini.core.php#ini.memory-limit
memory_limit = 2048M

[benjaminsmith]

Hi @ssoliveira,

Thanks for getting back so quickly. I have filed an internal bug report, #13692, and are currently working on the issue. While I don't have an ETA, we have maintenance releases for Nagios XI about every 3-6 weeks, and this should be taken care of in the next upgrade.

[sigmainformatique]

Hello @ssoliveira,

Same issue here (5000 hosts/50000 services + gearman), about 1 minute to display services in CCM.

Can you please tell us how many time does it take you to Apply a Configuration? ( about 1:40 minute here)

Regards,
Guillaume

[ssoliveira]

Hello @ssoliveira,

Same issue here (5000 hosts/50000 services + gearman), about 1 minute to display services in CCM.

Can you please tell us how many time does it take you to Apply a Configuration? ( about 1:40 minute here)

Regards,
Guillaume

Hi Guillaume,

I do not use Apply via dashboard.

We use some of the NagVis dashboards, which take some time out every time the services are restarted.

So we created a practice of just writing the settings during the day. And we put a routine in the crontab that validates if the settings are Ok, and in case we restart the nagios.

[ssoliveira]

Hi @ssoliveira,

Thanks for getting back so quickly. I have filed an internal bug report, #13692, and are currently working on the issue. While I don't have an ETA, we have maintenance releases for Nagios XI about every 3-6 weeks, and this should be taken care of in the next upgrade.

Hi Benjamin,

Thank you very much for your support.