Some graphs not graphing but realtime plots are working

[ctyler2621]

This is not your normal "Cacti not graphing" post. 99% of my graphs are working just fine. The only issue that I have is that Ubiquiti airMax radios are not working. Here are the specifics of the issue.

Cacti Version 1.1.10
SPINE 1.1.10

All other device types work perfectly.
SNMP queries from the Cacti server to a Ubiquiti device works.
Realtime plotting within Cactu plots the correct values so the OID and graph settings are correct.
The normal graphs are blank and the values underneath are all -nan but are created and do exist.

$ snmpget -v1 -mall -c'*********' 10.0.0.01 UBNT-AirMAX-MIB::ubntAirMaxCapacity.1
UBNT-AirMAX-MIB::ubntAirMaxCapacity.1 = INTEGER: 59

$ rrdtool last 3604.rrd
1498583402

$ rrdtool fetch 3604.rrd AVERAGE
ubnt_airMaxCapacity

1498497300: -nan
1498497600: -nan
1498497900: -nan
1498498200: -nan
1498498500: -nan
1498498800: -nan
1498499100: -nan
1498499400: -nan
(truncated, too many values to post here, but they are all the same)

What should I check next?

[ctyler2621]

Screen shot attached. Top shows the realtime graph and bottom is the same graph with realtime plotting turned off.

[Osiris]

More than likely file permissions. Check your Cacti log, or run the poller by hand as the crontab user. If you run by hand, you will see the permission denied messages. You can also get this by changing your cron entry to something like:

Code: Select all

*/5 * * * * apache php -q /var/www/html/cacti/poller.php >> /tmp/poller.log 2>&1

After a few poller runs, you can review for permission denied messages.

[ctyler2621]

I may have figured it out by accident. Looks like these all defaulted to the "High collection rate" profile instead of the "System Default" profile for some reason. As I don't have a poller set up to handle the high collection rate, they were failing to update. I'll give them a couple hours and see how it progresses now.

I was getting a lot of these "Set MAX_OIDS to 1 for this host to isolate bad OID" errors in my cacti log as well. Will update this post if this clears those errors as well.

-EDIT-
This turned out to the be one of the problems. The other was caused by using name OID's rather than the numerical OID's. Once those two items were "fixed" the graphs started populating properly.

[Osiris]

Good to know.

[frankvasquezf]

Hello,
I have the same problem but I can't fix it. I have only one graphic of disk space that not graphing but realtime is fine. I have 50 graphs of this type but only one is not working fine. Moreover, for the same host I have 2 graphs of that and only one don't work. Even I change the poller interval to 30 seconds and rebuild the poller cache, but I still with the same issue. I attached the screenshots.

My cacti version es 0.8.8h.

Thank you for your time and help.

[adma]

Hi frankvasquezf,

for me the issue was assert re-cache failed for data queries. After hitting "Verbose Query" for data queries under my device, everything started working again.

[ataha]

kindly explain with commands because i have same problem graph in real time only

[ataha]

in web page logs
POLLER: Poller[Main Poller] PID[170648] 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 61,438.2 seconds have passed since the last poll

[TheWitness]

You are cross posting. Please only post you topic once.