Cacti dont Graph every 15 minutes in a hour

drmspidi · Joined: 20 Jun 2008 Posts: 8

Does anyone know why in my graphs every hour, 15 minutes dont have any graphs ?? I've tried a lot of things but still this problem.

NeMiSiS · Joined: 25 Jun 2008 Posts: 14 Location: South Africa

Hi

First off, I dont know much about cacti .. I'm just thinking about something I saw similar in NeDI. have you tried timing the poller thread. Does the poller thread run longer than 15min? If so it might cause some data corruption/loss.

You can probably time it by loggin or su'ing to the user, running the poller.php/cmd.php with the time command in front and see how long it takes to complete.

drmspidi · Joined: 20 Jun 2008 Posts: 8

The poller is seted in 5 min. I dont think that this is the problem. But thanks any way !

TheWitness · Posted: Thu Jun 26, 2008 2:25 pm Post subject:

What does your cacti log say? Is this on a VM?

drmspidi · Joined: 20 Jun 2008 Posts: 8

He is giving me erros like that !!

07/01/2008 09:07:49 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (681, 'traffic_in', '2008-07-01 09:07:30', '159805355')'
07/01/2008 09:07:49 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (680, 'traffic_out', '2008-07-01 09:07:30', '0')'
07/01/2008 09:07:48 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (680, 'traffic_in', '2008-07-01 09:07:30', '0')'
07/01/2008 09:07:48 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (679, 'traffic_out', '2008-07-01 09:07:30', '3993428178')'
07/01/2008 09:07:48 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (679, 'traffic_in', '2008-07-01 09:07:30', '4051483468')'
07/01/2008 09:07:48 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (674, 'traffic_in', '2008-07-01 09:07:30', '242243544')'

and

07/01/2008 09:08:18 AM - CMDPHP: Poller[0] Host[40] DS[865] WARNING: Result from SNMP not valid. Partial Result: No Such Instance cur
07/01/2008 09:08:15 AM - CMDPHP: Poller[0] Host[35] DS[859] WARNING: Result from SNMP not valid. Partial Result: No Such Instance cur
07/01/2008 09:08:15 AM - CMDPHP: Poller[0] Host[35] DS[859] WARNING: Result from SNMP not valid. Partial Result: No Such Instance cur
07/01/2008 09:08:15 AM - POLLER: Poller[0] WARNING: Poller[0] Host[41] SNMP Time was not numeric

Any ideias ??

drmspidi · Joined: 20 Jun 2008 Posts: 8

This is another error.

07/01/2008 10:15:01 AM - POLLER: Poller[0] WARNING: Cron is out of sync with the Poller Interval! The Poller Interval is '300' seconds, with a maximum of a '300' second Cron, but 600 seconds have passed since the last poll!

gandalf · Posted: Tue Jul 01, 2008 3:16 pm Post subject:

Please check 2nd link of my sig and read the section "double poller issue"
Reinhard

drmspidi · Joined: 20 Jun 2008 Posts: 8

Sorry, i didnt undertand...

drmspidi · Joined: 20 Jun 2008 Posts: 8

07/02/2008 08:04:13 AM - POLLER: Poller[0] WARNING: Cron is out of sync with the Poller Interval! The Poller Interval is '300' seconds, with a maximum of a '300' second Cron, but 552 seconds have passed since the last poll!

What this means ?

gandalf · Posted: Wed Jul 02, 2008 9:29 am Post subject:

Misc: Debug NaN's in your graphs
at http://docs.cacti.net/node/358

pgb205 · Joined: 10 Nov 2008 Posts: 17

I'm getting the same sql 1062 errors and my machine is on the vm. It has a full gb of ram and dual core with plenty of hd space. I get the same choppy graphs as well.

TheWitness · Posted: Tue Nov 11, 2008 11:52 pm Post subject:

There errors are all about either a wandering clock, or dual cron's. I would guess that the guest machine has a large varying clock that is not being kept in check by VMWare or Xen, as often as it should be. You could force it using ntpdate in a 1 minute cron though.

TheWitness