cc , and took 1.21 seconds, Subject 'Cacti System Warning - Maximum runtime of 298 seconds exceeded for poller id 1. Exi

[eliassal]

Can somebody else with more expertise to fix or instruct on to find the root cause

[cigamit]

Please don't create duplicate posts, I have removed the other.

You can't just change your poller interval, as it affects your graphs and will cause issues. If you are collecting data at 1 minute intervals on your graphs, then I recommend both a 1 minute poller setting and a 1 minute cron setting, and change your actual cron to be 1 minute to match (and please don't change them again, you just break stuff if you do).

I would recommend you install spine and change from cmd.php to spine for polling.

[eliassal]

Thanks, I will change right now and let you know

[eliassal]

After updating all parameters according to your recommendation, I got exactly the same thing, but now it indicates 90seconds and not 298 seconds

[eliassal]

I also visited cacti download page there is no instructions on how to install on fedora, any hint or page on how to do this?
Thankks

[cigamit]

Yes, now install spine and swap to it as the poller.

[eliassal]

I installed cacti spine by
dnf install cacti-spine

updated etc/spine.conf
I visited settings page I see only cmd.php.
Anything else I should do or start something.....
Thanks

[cigamit]

On settings >> path, ensure the path to spine is correct. Then change the Poller Type to Spine in settings.

[eliassal]

I did and updated poller to use spine but still getting the same alerts when poller runs but when I run spine from command line on the cacti box it seems working fine

[eliassal]

I disabled the poller in settings and ran spine from command line and as I said it works without the alert and graphs started working

[eliassal]

I parameterized all 3 to 5 minutes, graphs getting updated but getting a message and alert indicating

2022-02-24 09:55:01 - POLLER: Poller[Main Poller] PID[2661394] WARNING: There are 1 processes detected as overrunning a polling cycle, please investigate
2022-02-24 09:18:01 - POLLER: Poller[Main Poller] PID[2537605] WARNING: There are 1 processes detected as overrunning a polling cycle, please investigate
2022-02-24 09:17:02 - POLLER: Poller[Main Poller] PID[2537454] WARNING: There are 1 processes detected as overrunning a polling cycle, please investigate
2022-02-24 09:16:02 - POLLER: Poller[Main Poller] PID[2537278] WARNING: There are 1 processes detected as overrunning a polling cycle, please investigate
2022-02-24 09:15:02 - POLLER: Poller[Main Poller] PID[2536941] WARNING: There are 1 processes detected as overrunning a polling cycle, please investigate
2022-02-24 09:14:03 - POLLER: Poller[Main Poller] PID[2536801] WARNING: There are 1 processes detected as overrunning a polling cycle, please investigate

How can I investigate in Cacti which process causing this?
Thanks

[eliassal]

Please help this is driving me crazy with the amout of email alerts I am receiving

[eliassal]

OK, after started to having graphs updated since I started using spine, all of a sudden, since 2 hours graphs are not updated, no alerts no emails devices are up and can be snmped !!!!!!

[eliassal]

Also, when I visite a device and click on refresgh associated data sources, I get the following errors in red in the log

2022-02-24 18:46:53 - CMDPHP PHP ERROR WARNING Backtrace: (/host.php[116]:host_reload_query(), /host.php[574]:run_data_query(), /lib/data_query.php[108]:query_script_host(), /lib/data_query.php[647]:CactiErrorHandler())
2022-02-24 18:46:53 - ERROR PHP WARNING: Undefined array key "script_path" in file: /usr/share/cacti/lib/data_query.php on line: 647
2022-02-24 18:46:53 - CMDPHP PHP ERROR WARNING Backtrace: (/host.php[116]:host_reload_query(), /host.php[574]:run_data_query(), /lib/data_query.php[108]:query_script_host(), /lib/data_query.php[647]:CactiErrorHandler())
2022-02-24 18:46:53 - ERROR PHP WARNING: Undefined array key "arg_index" in file: /usr/share/cacti/lib/data_query.php on line: 647

[cigamit]

Most likely you have a problem with something in one of the script templates. To debug, I would delete all graphs and then just create 1, and ensure things are processing properly. If not, then there is something strange going on with your system.