Script server: poller over ran its polling intervale

Post general support questions here that do not specifically fall into the Linux or Windows categories.

Moderators: Moderators, Developers

Post Reply
Author
Message
jmroth
Cacti User
Posts: 59
Joined: Sun Jan 20, 2002 7:00 pm

Script server: poller over ran its polling intervale

#1 Post by jmroth » Wed Mar 20, 2019 3:53 am

Hi,

Ever since implementing a script server query (indexed) script (using SNMP) to retrieve the data from viewtopic.php?f=21&t=59681 I am getting the following:

Code: Select all

2019/03/20 08:30:02 - CMDPHP WARNING: cmd.php poller over ran its polling intervale and therefore ending
2019/03/20 08:30:02 - AUTH LOGIN: User 'admin' Authenticated
2019/03/20 08:30:06 - POLLER: Poller[1] WARNING: Invalid Response(s), Errors[4] Device[8] Thread[1] DS[1330, 1509, 1510, 1511]
2019/03/20 08:30:09 - POLLER: Poller[1] WARNING: Invalid Response(s), Errors[3] Device[9] Thread[1] DS[1500, 1501, 1502]
2019/03/20 08:30:14 - POLLER: Poller[1] WARNING: Invalid Response(s), Errors[3] Device[10] Thread[1] DS[1503, 1504, 1505]
2019/03/20 08:30:16 - POLLER: Poller[1] WARNING: Invalid Response(s), Errors[3] Device[11] Thread[1] DS[1506, 1507, 1508]
2019/03/20 08:34:26 - POLLER: Poller[1] WARNING: Invalid Response(s), Errors[1] Device[46] Thread[1] DS[5097]
2019/03/20 08:35:01 - POLLER: Poller[1] Maximum runtime of 298 seconds exceeded. Exiting.
In order to solve the issue from the other thread, I guess using a data query is not possible, so what are my options here in order for performance not to suffer? (I thought a script server query was the most effective one?)

Thanks.

NOTE Current stats:
Hosts:136 HostsPerProcess:136 DataSources:6957 RRDsProcessed:3557

When we add 100 graphs using said script query, then the poller is dead (see above)


PS. Tried searching the forums BTW: "The following words in your search query were ignored because they are too common words: script server." Great

netniV
Cacti Guru User
Posts: 2872
Joined: Sun Aug 27, 2017 12:05 am

Re: Script server: poller over ran its polling intervale

#2 Post by netniV » Wed Mar 20, 2019 6:00 am

Yeah, the forum search can be annoying. The above error comes about because it takes too long to process all the data sources required for the graphs that need to be updated.

If you want to see the difference, maybe remove the graphs, monitor the polling time. Add ten, monitor the polling time. If there is a significant increase, you either have a slow responding device with too much data, or your script is just taking too long to complete.

For devices that may overrun the 1 minute polling time, try putting them into the 5 minute data collection profile. You should remove and add the graphs again if you are doing that though as simply changing them won't correct the RRD's.

jmroth
Cacti User
Posts: 59
Joined: Sun Jan 20, 2002 7:00 pm

Re: Script server: poller over ran its polling intervale

#3 Post by jmroth » Wed Mar 20, 2019 12:49 pm

I see. Ehm, what is the order of things how script server queries are called anyway? I tried to find out empirically, and it seems to need 'index', 'query', 'get', ... Is that documented somewhere?

It seems to 'get' each single value, why can't it just walk a subtree?

Also, can you have global variables inside your "ss_...." script (meaning data accessible for the entire execution of the poller run), or exactly how is this instantiated? (Okay this part is answered: I forgot to use "global" since all the stuff is in functions...)

Thanks.


jmroth
Cacti User
Posts: 59
Joined: Sun Jan 20, 2002 7:00 pm

Re: Script server: poller over ran its polling intervale

#5 Post by jmroth » Thu Mar 21, 2019 6:36 am

I guess the point is that it has to be looked at together with Data-Queries.md doc (which does not even mention Script server).

Well if I find some time I'll see if that can be streamlined. :)

Post Reply