I cannot even find a way to create port groups. Searched Google, looked all over my installation and cannot find such an option.
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Tuesday, September 27, 2016 10:51 AM To: observium@observium.org Subject: Re: [Observium] Trunk Issues
Why wouldn't it?
I think you're perhaps expecting the "total device traffic" graph to be something specific to your use case, rather than, well, the total device traffic.
You probably want to create a port group and look at that instead!
adam.
On 27/09/2016 05:58:23, Talk Jesus <chad@talkjesus.com mailto:chad@talkjesus.com > wrote:
Problem:
Why is Observium calculating the “uplink” to the trunked 2nd switch in the total utilization graph? This is throwing off accuracy.
How do I add the 2nd switch as a device so it’s interfaces show up too. Right now, only the primary switches’ interfaces are there
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Monday, September 26, 2016 10:51 PM To: Observium Network Observation System <observium@observium.org mailto:observium@observium.org > Cc: Observium Network Observation System <observium@observium.org mailto:observium@observium.org > Subject: Re: [Observium] Trunk Issues
I'm not really sure what the problem is that you're trying to describe.
Adam.
Sent from BlueMail http://www.bluemail.me/r
On 26 Sep 2016, at 21:06, Talk Jesus <chad@talkjesus.com mailto:chad@talkjesus.com > wrote:
I've been using Observium for a while and there seems to be some discrepancies between my graphs and the upstream (my data center provides me Ethernet GigE bandwidth). For example, from around 8am to 12pm today it showed overall utilization pretty consistent well above 400Mbps. I have two drops from them: primary and secondary (redundant). Plus, they both show active instead of active/passive. Not sure if that has anything to do with it.
From the tech:
"95th percentile is calculated from both ingress and egress traffic. For example, if you used a constant 200Mbps up and down, then you would only be charged your commit of 200Mbps, not double. I do see what you mean that the two handoffs (layer 2) are equaling more than 300Mbps, but when you look at the Aggregate (layer 3) graph at the top, it does not go over 300Mbps. These packets can possibly be queued before going out so you don’t go over your commit. "
So, what do you think might be causing this hiccup?
I noticed this too: so these switches are actually trunked (2 x 3650 L3 switches). In the graph, it also shows the uplink connectivity calculated with the total utilization. This is definitely not accurate as it shouldn't do that.
Example:
Drop #1 = 40Mbps
Drop #2 = 200Mbps
Uplink to Switch #2 = 130Mbps
Now as a result VLAN2155 (total utilization) is calculating not only #1 and #2 drops from the upstream, but even the trunk port labeled "Uplink to Switch #2".
How can I rectify this with Observium to avoid that and get an accurate output?
_____
observium mailing list observium@observium.org mailto:observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium