SNMP Causes Graph Spikes after Reboot

[thesg]

Is there a way without having to use killspike.pl to get rid of the spikes in my graphs when a device is rebooted?

I checked some of the settings but wasn't sure if I put a max it would have an effect when something went off the scales (like after a reboot).

thanks..

[thesg]

How is everyone else dealing with these spikes?

I only see this with Cisco devices, btw.

[BSOD2600]

Apparently, the next version of cacti will have some sort of automatic fix for this. Otherwise, just need to run the killspike.pl script in the mean time.

[bulek]

Proper maximum value = no spikes. Any value higher than max is just dropped (because it is incorrect from your point of view). Otherwise you have to stick with killspike.pl.

- Piotr

[crinago]

Proper maximum value = no spikes. Any value higher than max is just dropped (because it is incorrect from your point of view). Otherwise you have to stick with killspike.pl.

- Piotr

How do I set the maximum value?????

[bulek]

You can configure it in the data source settings. If your RRD file already exists you need to do additionally one of two things:
- delete the file and let Cacti recreate it from scratch
- use 'rrdtool tune' command to update maximum value inside a file

- Piotr

[PKLNTKJ334]

where do I get killspike.pl script???

[bulek]

http://people.ee.ethz.ch/~oetiker/webto ... kes.tar.gz