-
Notifications
You must be signed in to change notification settings - Fork 161
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Aggregate MRTG graphs are including the reseller uplink interfaces as well as the peering interfaces, creating incorrect graphs/stats. #855
Comments
is the reseller NNI physical interface set to type: |
@ichilton - I'll need blow away some cobwebs to look at that code - hasn't been changed in years and it seems slightly odd that this hasn't come up before. E.g. if this is the cause, graphing the reseller port would be done in both fan out mode and veth hairpinning mode, I think. (edited to remove reference to disabling polling - that's per-switch, not per-interface) |
Yep! ...and the two sub-interfaces set to 'Peering'. |
As I say, we only noticed it because this 'reseller' is doing more than a trivial amount of traffic (40Gbps in this case). Happy to send screenshots or screen share demo on a zoom call if you think we might have something slightly different to you / wrong, rather than the code? ...but the parent interfaces are definitely set to type 'Reseller' and are appearing in the MRTG aggregated graphs. Thanks, Ian |
Fixed via e40be75 |
ISSUE TYPE
Bug Report
OS
VERSION
(latest on release-v6 branch)
ENVIRONMENT
CONFIGURATION
Don't want to put that on GitHub.
SUMMARY
We are in the process of moving from physical loopback cables to EOS layer 2 subinterfaces to provide reseller ports.
We have just his morning moved our first which is doing a significant amount of traffic (~50G) as it's actually an ASN share rather than reseller.
We are now seeing that our traffic graphs have ~40G difference between IN and OUT traffic.
I had a hunch what this would be and have confirmed it - the MRTG aggregate graphs are graphing the reseller uplink interfaces as well as the peering interfaces.
i.e it's graphing both the traffic on the physical port as well as the subinterfaces.
I have verified this by manually editing mrtg.cfg and removing each occurrence of that physical port and the IN and OUT have now converged to within a few hundred Mbps, which is about right for the other resellers.
It should just be including the 'peering' ports, not those set to 'reseller'.
IMPORTANCE
HIGH - this is affecting our production traffic graphs.
The text was updated successfully, but these errors were encountered: