What's new

Traffic Stats - Which to believe?

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

CaptnDanLKW

Senior Member
I love my data and stats but I'm not sure which to believe.
The Data in Traffic Monitor (Merlin's Enhancement) or Traffic Analyzer - Statistics (Part of Trends DPI engine)

Traffic Analyzer:
upload_2020-4-30_7-15-25.png

or Traffic Analyzer - Statistics:
upload_2020-4-30_7-16-45.png


Router is an RT-AC86U on 384.17. I have QOS Disabled, so Runner and FA acceleration are enabled. I remember from the AC68U and N66U days that the CTF / HW Acceleration could skew the stats in Traffic Monitor because the bits bypass iptables... or something like that.

So, which do I believe?

Dan
 
Neither. :)
 
Compare with your ISP data, you'll see.

At first glance, your Traffic Monitor data is twice that of your Traffic Analyzer.
 
I love my data and stats but I'm not sure which to believe.
The Data in Traffic Monitor (Merlin's Enhancement) or Traffic Analyzer - Statistics (Part of Trends DPI engine)

Traffic Analyzer:
View attachment 23165
or Traffic Analyzer - Statistics:
View attachment 23166

Router is an RT-AC86U on 384.17. I have QOS Disabled, so Runner and FA acceleration are enabled. I remember from the AC68U and N66U days that the CTF / HW Acceleration could skew the stats in Traffic Monitor because the bits bypass iptables... or something like that.

So, which do I believe?

Dan

Did you ever get to the bottom of this, Dan?

I've just noticed the same is true for me, using July 1-8 as the comparison (we've been away from the house, so data usage is low):

Monitor.PNG


And the same in analyzer:

Analyzer.PNG


More or less double. Not that I trust my ISPs data monitoring (8.93GB at time of writing, but won't have accounted for today's usage yet), but it's a lot closer to the Analyzer numbers, than the Monitor numbers. For reference in June 2020, Monitor showed 573GB, Analyzer showed ~518GB, and my ISP showed 439GB.

I have QoS enabled, set as adaptive, and strangely HW acceleration is enabled for both runner, and flow cache - i assumed having QoS enabled, would set those to disabled.

Thanks!
 
In short - no. I never got a clear answer in this thread nor a lot of other searching. I think the variables (models, QOS types, HW accel (two types), per-IP monitoring, make it difficult for a single clear answer why, I just wish I knew how things worked under the hood so I could do real time checking.

That said, as of 6/1 I decided to start over with all my trend data (did not re-init my tomato stats). Turned off all the trend stuff (shut everything off, removed all the .db files), and re-enabled. That was with 384.17 (and now 384.18). and my Traffic Status / Traffic Analyzer values about the same... +/-.

No other changes - Still an AC86U Primary Mesh + 1 AC86U Node, QOS has been off, per-ip monitoring off. That means Runner and FC HW accel are both enabled.

TA:
upload_2020-7-8_21-29-7.png


TM:
upload_2020-7-8_21-29-51.png



So I don't know what happened or why things have evened out but I'll keep my eye on it.

My fuzzy memory tells me that per-IP monitoring and 'traditional' QOS are the two likely culprits that usually cause inaccurate counters but YMMV.
 
per-IP monitoring is not present on the HND routers as it is incompatible.

Well I'll be darned - just assumed it was there and I never enabled it. Observation noted. +1
 

Similar threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top