Cacti

We recently upgraded to 087e with spine and are having issues with 2 pollers running. I know there is only one, because when I comment the line out of my /etc/crontab I get nothing. Any help would be greatly appreciated. 05/07/2010 11:21:39 AM - SYSTEM STATS: Time:157.6946 Method:spine Processes:2 T...

If you go into the Graphs tab > Settings on the far right > you should see Page Refresh.

It should say 300.

I had a similar issue a while back and it turned out to be yet another Firefox update that broke my refresh.

I personally have never seen a poller that high. I've only seen 1 minute and 5 minute. What happens when you poll every 5 minutes? Does the device show down or just doesn't grab any information. I'm not familiar with the templates you are using, but is it possible to glean the mib's and try to walk ...

I'm almost certain that the issue is the file needs to be resized and it is dropping data as it goes. From the time I posted yesterday until today, another day has dropped off of the historical graph.

Is there some tricks or debug that I can run to troubleshoot this?

Thanks

I am having a weird issue with RRD's. For some reason I'm only maintaining about 5 months of graphs for viewing "past year". Here are my sample settings: /usr/bin/rrdtool create \ /var/www/html/rra/406/29248.rrd \ --step 60 \ DS:traffic_in:COUNTER:600:0:1000000000 \ DS:traffic_out:COUNTER:...

We found a template here in the forums called "Cisco Gateway Active Calls" and it has been spectacular. However... We are having an issue providing provisioning and planning information since Cacti is "normalizing" the data with the other graphs. Is there a way to keep this data ...

I am doing the same thing with many different routers using "Cisco Gateway Active Calls". I think I found it here in the forums. It plots how many concurrent calls I have on each gateway. Using thold, I can then send an alert when we come close to maxing out. The only issue I have is that ...

How did you install rrdtool and how did you install Cacti?

I am querying 3k switches/ routers using a RO community without issue.

I had a few issues starting out where I had some bad copy/paste when adding devices.

Are you manually typing the community in or pasting from the switch config?

have you verified that opt/lampp/htdocs/cacti/rra exists? Have you also checked the permissions of that directory?

Is there anything in the cacti.log relating to this?

We recently replaced some 4500's with 3750 stacks throughout our campus. Cacti is still showing the old interfaces (g2/4) and the new ones (g2/0/4). Is there a way to delete these erroneous data_sources/ graphs easily? I tried to use rrdcleaner, but it appear to only find files that do not have DS's...

yum install dejavu-lgc-fonts.noarch

This worked for me.

I know how to add interfaces via the cli and add_graphs.php, but was wondering if someone had some slick way of purging old interfaces. An example would be a 48 port card removed from a 6500. Obviously the ports would no longer be there, so cacti should be receiving an error when polling those mibs....

this would be awesome if you could get most of the info. We have been trying to accomplish in our spare time little results.

How did you install cacti? Source, rpm, or other?

Do you have to pollers running?