Shaw → CGNM-2250 Problems (CGNM-2250 VS DPC3825)

Ok, Shaw called me and offered me 120, i wasn't sure about it, so they offered to give me 120 and let me keep my 100 while i choose, offering to refund the difference when i do.

I have internet 120 on a CGNM-2250 (only modem aloud with this plan) and internet 100 on a DPC-3825 (the only modem that can be bridged with internet 100)

It should also be noted that both modems are bridged, they are not doing any routing, they are basically just acting as a dumb switch, when bridged the only thing the devices are required to do is be a modem, thus i feel the problems with the CGNM-2250 are with poor modem components in order to bring costs down, unfortunately the performance drop this creates is unacceptable to me.

I have both modems bridged connected to my Ubiquiti EdgeRouter Lite, the router is bonding the connections, they both have completely equal priority, I'm getting about 180/12 wired, simply because bonding with my router uses up a lot of CPU so i cant get full download, but i do get 128 on the 120 and 107 on 100 which would be 235 if i had the CPU for it.

I would image a firmware fix could resolve this issue.

Unfortunately, I suspect that its more than that, although one would hope not. The Rogers CGN3, which predates the 2250 by a few months is just as bad. There was a CGN3 firmware update recently, which seems to have brought the max ping times to the modem down from 100 ms to the 20 ms range. Lets see, one would expect 1 ms from a router or less, hmm. Not looking good at this point. So, either its a case of the Hitron modems being woefully underpowered, in terms of their processing capability, or the firmware is in serious need of review and overhaul, take your pick. If all of the Canadian ISPs are headed towards Hitron modems, there will be no choice for Canadian consumers looking for high speeds and reasonable performance, which includes latency. Fwiw.

These tests are being run concurrently?
If so, have you tried smoke pinging them individually?

they are concurrent, and yes i have done individual, same results. infact i started only running it on the hitron and added the DPC just to show how shit the hitron is

upgraded to an edgerouter pro today, getting closer to that 235/12 max i was hoping for

i dont have any limiters on downstream but i do on upstream, here is pings to google while my connection is under full load

Reply from 216.58.216.174: bytes=32 time=16ms TTL=57
Reply from 216.58.216.174: bytes=32 time=14ms TTL=57
Reply from 216.58.216.174: bytes=32 time=41ms TTL=57 - download test start
Reply from 216.58.216.174: bytes=32 time=52ms TTL=57
Reply from 216.58.216.174: bytes=32 time=93ms TTL=57
Reply from 216.58.216.174: bytes=32 time=89ms TTL=57
Reply from 216.58.216.174: bytes=32 time=58ms TTL=57
Reply from 216.58.216.174: bytes=32 time=51ms TTL=57
Reply from 216.58.216.174: bytes=32 time=79ms TTL=57
Reply from 216.58.216.174: bytes=32 time=58ms TTL=57
Reply from 216.58.216.174: bytes=32 time=59ms TTL=57
Reply from 216.58.216.174: bytes=32 time=65ms TTL=57
Reply from 216.58.216.174: bytes=32 time=64ms TTL=57
Reply from 216.58.216.174: bytes=32 time=51ms TTL=57
Reply from 216.58.216.174: bytes=32 time=39ms TTL=57
Reply from 216.58.216.174: bytes=32 time=52ms TTL=57
Reply from 216.58.216.174: bytes=32 time=84ms TTL=57
Reply from 216.58.216.174: bytes=32 time=42ms TTL=57
Reply from 216.58.216.174: bytes=32 time=21ms TTL=57
Reply from 216.58.216.174: bytes=32 time=16ms TTL=57 - end
Reply from 216.58.216.174: bytes=32 time=14ms TTL=57
Reply from 216.58.216.174: bytes=32 time=16ms TTL=57
Reply from 216.58.216.174: bytes=32 time=16ms TTL=57
Reply from 216.58.216.174: bytes=32 time=16ms TTL=57
Reply from 216.58.216.174: bytes=32 time=17ms TTL=57 - upload test start
Reply from 216.58.216.174: bytes=32 time=17ms TTL=57
Reply from 216.58.216.174: bytes=32 time=20ms TTL=57
Reply from 216.58.216.174: bytes=32 time=21ms TTL=57
Reply from 216.58.216.174: bytes=32 time=19ms TTL=57
Reply from 216.58.216.174: bytes=32 time=17ms TTL=57
Reply from 216.58.216.174: bytes=32 time=24ms TTL=57
Reply from 216.58.216.174: bytes=32 time=18ms TTL=57
Reply from 216.58.216.174: bytes=32 time=22ms TTL=57
Reply from 216.58.216.174: bytes=32 time=21ms TTL=57
Reply from 216.58.216.174: bytes=32 time=22ms TTL=57
Reply from 216.58.216.174: bytes=32 time=23ms TTL=57
Reply from 216.58.216.174: bytes=32 time=22ms TTL=57
Reply from 216.58.216.174: bytes=32 time=22ms TTL=57
Reply from 216.58.216.174: bytes=32 time=22ms TTL=57
Reply from 216.58.216.174: bytes=32 time=26ms TTL=57
Reply from 216.58.216.174: bytes=32 time=23ms TTL=57
Reply from 216.58.216.174: bytes=32 time=14ms TTL=57 - end
Reply from 216.58.216.174: bytes=32 time=16ms TTL=57
Reply from 216.58.216.174: bytes=32 time=15ms TTL=57
Reply from 216.58.216.174: bytes=32 time=15ms TTL=57

whats the cpu usage at with that.

spikes at 11% when i run a test, could handle two bb250 lines or 2 telus 100 lines easy

Well, it is ridiculously new. Hopefully there will be a firmware fix in the near future. I was really hoping this would be a kick-ass router that would satisfy the hardcore crowd such as yourself.

Swing and a miss.... Unless they can tweak it.

ya im very disapointed about the CGNM-2250, all we can do is complain at this point, when i posted on the shaw forums no one seemed to care, and the lack of people that care about this is still very high

Yeah, but most of you in here are power users. Yourself maybe more so than the rest. The majority of folks won't be bothered by a few ms difference.

Question: is anything less than 40-50 ms (total) really going to make a difference for gaming or mission critical applications?

Was it fairly easy to setup the edgerouter like you have?

yes, very easy. but thats not for this thread, pm me if curious plz

My CGNM-2250 in bridge mode has gotten stuck twice so far.
When it happens, I can't ping anything.
Resetting my router did not fix the problem, but resetting the CGNM-2250 did solve it.
This points to some kind of firmware issue with the CGNM-2250.

I have also noticed quite variable pings on the CGNM-2250 (high jitter.)

The 120 plan is decent, but I am not satisfied with this modem.

yes, high jitter is the issue, if the ping were 1-5ms higher whatever, but that jitter is what pisses me off.

Why can't you have 120 on the Cisco modem? Or can you if you have it all ready?

About to move and just praying I can keep BB250. Different node same city so.... I know lots of people still on 100/10 and 50/5 in this city.

Not sure what I would do if I had to go down to 120 from 250 for basically the same price.

anywhere 100/10 or 50/5 is avail bb250 is, places which are limited to 100/5 and 50/3 are the places 250 aint avail, but some places where 250 is avail have been reverted back to 100/5 50/3 only

The place I work is literally in the middle between my current and new house. We have 50/5 at work but could not get 100/10 or 250/15 at work.

Going to move Shaw stuff myself and see what happens. Worst case I know the Shaw construction manager from where I work and I'll bug him with questions. (and make him bury our second line!!!!)

Will update this after March 13 when I move in.

Fryst,

The modem isn't the issue, but the firmware. Hitron is already aware of the issue and working on new firmware with fix. Once the firmware is deployed, it will address the issue.

Yeah, Business 50 used to be 5 upload everywhere

They changed it to 3 sometime last year.

The Hitron deprioritizes ICMP traffic causing any ICMP echo (ping) testing to be highly inaccurate.

Ookla speed tests (speedtest.net, speed.shaw.ca) use TCP based latency testing. You will noticed those tests show an accurately low latency.

So the ICMP numbers you see are artificially high, the actual latency for TCP data flow is typical 8ms - 12ms for the first Shaw hops. Unfortunately so many things (network performance testing, online gaming) depend on the ICMP latency figures, so even though the TCP latency is nice and low, the poor handling of ICMP (and maybe UDP) traffic needs to be solved.

Shaw engineering is aware and the vendor is working for a firmware resolution.

As some background, I'm on a highly saturated node that is scheduled for a node split to be completed by the summer. Around 8am-9pm I see less than 3mbit download/upload regardless. My Cisco was bonding 8x4 channels, the Hitron is bonding 12 downstream (the max available in my area), upstream it's bonding more than 4 as upload is higher now, however I didn't ask engineering how many upstream channels were in my area.

But here are some of my tests comparing the Cisco DPC3825 (bridged) and the Hitron CGNM-2250 (Pass through, no Bridge code for BB250 yet). These were selected to represent close usage periods, not to skew anything.

Non-peak hours on Hitron

good info, this does make sense as i get similar ping results to speedtest but never icmp, very good post!

interesting to here how many upstream channels r on ur node? or maybe even on mine? hmmm

@predator314Are you able to indicate where this interesting statement comes from and how it was determined? I run a Hitron CGN3, which is the version prior to the CGNM and CGNM-2250 used by Rogers and now Shaw, but the apparent latency is still the same problem. This is a common complaint on the Rogers forums.