[ixpmanager] Spikes on MRTG graphs

[Martin J. Levy]

This is (at least with the second graph) a known issue with MRTG systems. If I recall correctly you should be able to edit the raw data file (in-between 300 second polling intervals) and replace the value from its obviously too-high number back down to a reasonable number (I.e. something averaging tge before/after numbers). His type of MRTG hack has to be documented somewhere - but I don’t know where.

There’s also some max limits that can be set in your config file so that you don’t see this bad a spike in the future.

Remember - MRTG data file are fixed length, so don’t just delete a line. Oh - and back your file first. Plus - test editing first by using a spare port and it’s data file. 

Martin

> On Mar 27, 2020, at 6:05 PM, Vladislav Bidikov <vladislav.bidikov at finki.ukim.mk> wrote:
> 
> 
> Hi all you IXP people,
> 
> We seem to have some strange issue with spikes happening when our 2 switches get the connection back (after a fiber was cut)
> Although this looks cool like a very good IXP in reality it destroys our graph levels...
> 
> http://prntscr.com/ro54fc
> 
> http://prntscr.com/ro54kh
> 
> 
> Any idea how to fix this or maybe smooth it...
> 
> IXP.mk is getting momentum and this is still a issue we see at least two time this week (we hope there will be no more fiber cuts in the next few months)
> 
> On the other hand - any good tutorial on how to setup Sflow with Quanta QNOS and Linux ?
> 
> 
> Best regards,
> 
> -- 
> V.B
> FCC
> 
> _______________________________________________
> INEX IXP Manager mailing list
> ixpmanager at inex.ie
> Unsubscribe or change options here: https://www.inex.ie/mailman/listinfo/ixpmanager
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.inex.ie/pipermail/ixpmanager/attachments/20200327/59fe5c79/attachment.htm>