Cacti stops graphing

[yuji]

Hi,

For the past several months I have been using Cacti, Cacti has been working great. I recently just had a major network upgrade and for awhile it worked seamlessly until now. I was attempting to create image files on the local directory and cacti suddenly wasn't working for some reason. It had major gaps in its polling. I decided to remove graph exports and delete the directory where it placed these graphs but this did not resolve the issue. I tried rebooting the linux server and now it does not seem to poll at all, including localhost. I made sure all the settings are correct. The poller is in the crontab file and mysql, httpd, and poller seems to be running when I type ps -ef. Is there anything else I am missing?

Yuji

[yuji]

Also, cacti is not able to poll itself as well.

[gandalf]

Please switch Logging Level to DEBUG and have a look at log/cacti.log. As it seems to be an intermittend error, you may have to run this for a couple of cycles.
Make sure, that the poller is started from one crontab only (/etc/crontab, /etc/cron.d/cacti, crontab -e -u root, crontab -e -u cactiuser, ...)
Reinhard

[yuji]

I am only using crontab for starting the polling. I ran the debug mode but can only do it for a short time until my system freezes. It polls the nodes fine and places it into the mysql table

"02/09/2006 03:53:05 PM - CACTID: Poller[0] Host[142] SNMP: v2: 172.16.1.14, dsname: traffic_in, oid: .1.3.6.1.2.1.2.2.1.10.10, value: 2387247957"

"SQLCMD: insert into poller_output (local_data_id,rrd_name,time,output) values (1349,'traffic_in','2006-02-09 15:45:08','2387247957')"

However I am still not seeing the graphs. The devices seem to be reached fine. The device list shows up fine and shows snmp information. I do a SNMP interface Verbose query and it retrieves the snmp information.

Anything else that you think might be the case? None of the device, including cacti, shows up on the graphs.

[gandalf]

You may change your crontab entry to

Code: Select all

*/5 * * * *     cactiuser       /usr/bin/php /var/www/html/cacti/poller.php > /var/www/html/cacti/log/poller.log 2>&1

Soo you'll find the poller's output in log/poller.log (for the last run only, so don't worry about disk space). There, you'll may find rrdtool errors.
Another approach: find rrdtool update statements in log/cacti.log for the failing device when running in DEBUG mode. "su" to cactiuser and run those update statements from cli to see, if they fail (perhaps you'll need to adjust the timestamp, rrdtool requires 1 sec difference at minimum).
Next approach: run rrdtool fetch <rrd file> AVERAGE to verify, that these files get updated
Reinhard