Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Chrysalis]

Well PPP has caused some complaints in the opensource community, usually the pattern is someone complains of slow speed for the hardware, once its revealed they on PPP that gets blamed, its considered to be an inefficient, heavy on CPU, unable to be multithreaded and such.  If its heavy on consumer devices it may also be on the ISP side, hence the gains for cityfibre?

[burakkucat]

I'm that guy that doesn't even run BQM as I don't see the point. Wonder if I'm the only anything like regular user that doesn't have CQM or BQM?

You are not alone. A certain grumpy old black cat shares the same view and has never used BQM or CQM.

As for the excentric who started that fad . . . 

[Alex Atkin UK]

Its useful to be able to look at if you are having problems, or to confirm QoS is working.

Though I just discovered pfSense does keep records of its gateway monitoring which is interesting.

Though I wish pfSense was more clever about its monitoring as it relies on the gateway responding to ping, not all ISPs do that.  Would make more sense if it did PPP echos for PPP WANs, or monitored the quality of actual packet loss on the links themselves rather than relying on lost ping packets to detect a problem on the WAN.

I suspect its this relying on ping that is largely my problem with instability as it makes the mobile networks look less stable than perhaps they are, due to ICMP probably being low priority.

[Weaver]

I agree with the earlier posters, I never ever look at CQM unless there’s a problem. I don’t obsess over it. I do look at it to see what’s been happening with traffic into/out of my LAN though, when I’m unsure about something. But I disagree with my dear friend Burakkucat, in that I do find it very useful indeed at times, I just don’t obsess about it.

[Chrysalis]

In the I run it all the time camp but dont obsess.

If I had a problem with my broadband and then had nothing monitoring it to help diagnose it would be annoying.

[Alex Atkin UK]

I'm sure I have no idea what you mean about obsessing over stats. 

[Chrysalis]

Nice, what UI is that from?

[Alex Atkin UK]

Based on my interpretation of the Star Trek LCARS interface.

Lots of fudgery in the background with PHP scripts polling SNMP, directly from the server itself, plus pulling stats from more PHP scripts running on pfSense.

I have my server boot straight into a full-screen window in Firefox to render it and I have 10" touch-screens in two different rooms hooked up to it.  So I get a summary of the network and the server itself at a glance.

To keep things on topic, I might see if I can put WAN latency charts in that space underneath once I figure out how to extract the data from pfSense, as I do tend to login to pfSense a lot to look at that data when the mobile WANs are being glitchy.

[Jon21]

Had a response from A&A:

Sorry for the delay. I suspect location here is the key. Our CQM will be more accurate because the Firebricks that monitor your connection are on the same network. If you're still seeing the latency spikes on Think BB graphs then maybe try pinging 81.187.81.187 and see if you see the same increase.

I've sent them a link to the BQM that is monitoring that IP address.

[Alex Atkin UK]

Will be interesting to hear what they have to say about it.

[Jon21]

Will be interesting to hear what they have to say about it.

Still don’t think they are that bothered by it, despite giving them the BQM for the gateway. Do have a review date of Friday but I’m not sure if that is for the latency or whether it’s to do with the interleaving.

[Alex Atkin UK]

But if its not an issue, they might at least be able to explain why.  They are the only ones who can really look into the routing and see if its a real-world problem or not.

Considering BQM is also using their hardware, they may also have an idea if there is a glitch at TTB.

[Jon21]

You can ask them to make sure you are on the "Speed" DLM policy.
The Standard policy only allows half the number of ES before applying interleaving.

This was spot on. On the standard policy, looking to get it changed to speed, sometime next week.

[Jon21]

Well something has changed just after 9pm.

[Chrysalis]

Thats interesting, I had the same on my graphs , but then after midnight it went back to normal.