Read the forum guidelines
For about one year now, I've had an Arris SB8200 cablemodem and Cox 300Mbps Internet tier, and this has worked esssentially flawlessly in that time. I've had Cox Internet even longer with very few issues - until about two days ago.
I noticed decreased download speeds, and when I logged into my account, I was notified that an "issue" in my area was detected, and I assumed my speed issue would be resolved when the issue was solved. About two hours later, the issue was clear, but my Internet speeds were still decidedly abnormal. I then noticed that the middle-two lights on the front panel of the cablemodem indicated the *downstream* connection was DOCSIS 3.1 (bonded downstream), colored blue, but the *upstream* light was now *green*, indicating only DOCSIS 3.0. This is a change from the previous behavior in which both up and down lights were bonded (blue).
That was two days ago. The problem has persisted, across multiple reboots of my cablemodem, and reboots of the firewall appliance that receives the Ethernet out from the cablemodem. I contacted Cox and they said the signal looked clear, but they agreed that my struggle to reach 200Mbps (usually topping out around 170-190) is not consistent with the 300Mbps service tier. They went so far as to re-provision the modem, all to no avail.
I took a look at my cablemodem serivce page, and it shows that the upstream channels hover between 39.0 and 42 dBmV. I found that Arris states the required upstream signal must be between 45-55 dBmV.
A bit of Googling has revealed a bit of warfare normally has to be undertaken to get Cox to increase power to the upstream channels, which is precisely what the Arris tech support rep told me when I reached out to them for assistance (cablemodem is only one year old and under warranty). He flatly stated the ISP must increase channel strength. Excerpt of my upstream channel status listed below:
I traced the line all the way back to the Cox input, eliminating an intermediate cable in the attic, and directly attached the Cox coax to the cablemodem and powered it up - no difference - bonded down, unbonded up. This means the signal to the modem from Cox is too low. But if I know from research Arris says Cox has to increase power, but Cox will refuse, what on earth do I do? Go out and drop another $200 on a new/different cablemodem?
Any suggestions appreciated.
Your service was "esssentially flawlessly" on 11 May, so everything within your demarc is fine. After the issue in your area, your upstream seems to have not fully bonded because I only see 5 channels as opposed to 8. Try logging into your Cox account to reset your connection.
I agree, and I've told this to each of the four different people I've "chatted" with AND the level 2 tech who talked to me on the phone the other day. I've made a point of saying this seemed to have started with an outage affectingy area. Yet the last word from Cox is for a truck roll.
This is another aspect of my frustration: When I mentioned this to the first agent with whom I chatted he told me he would escalate it because of the outage being coincident with my issue. I waited for a couple of hours and checked back, told the next tech the situation, and they said there was no record of any tech escalating anything on my account to anyone for any reason. So I was lied to to get someone's chat resolution metrics inline. And I still have the problem.
I chatted with, as I recall, four different agents, got disconnected twice, and then told I had to pay for more diagnosis. That's when I about threw my phone through a wall. Then I called tech support and they related me to a (supposedly) Tier 2 who finally concluded a truck roll was necessary because everything in their end looked correct. I told them about the outage but it made no difference to them.
I'm just relaying what the tech observed. The cable run in the attic is basically in a summer roaster and after, what, twenty years (?) the heat maybe has gotten to it and the impedance finally just hit the "straw that broke the camel's back" moment. I don't have a hard answer at this point. I don't know why the Cox tech from years ago put a splitter on that line, but he did. We will see if the new run fixes the issue.
Your attic must be some roaster. Even the cheapest coax can survive 175º and probably more since it's not dry-rotting in sunlight.
After the tech yanks it, I'd check the cable for any obvious chaffing or severe kinks. If so, yeah, probably a bad cable. If not and the tech also suspects heat, will the tech replace with higher-quality (Ethylene, Propylene, PVC) or just what's in the truck?
I have no insight into what media he will use to run the new line.
You shouldn't know. Just questions for you to ask the tech. Do you subscribe to Complete Care? If not, what happens if new cable doesn't fix problem...no charge?
I does get too hot in attics for you, I and stocky guy...but not for metal and hard plastic. No electrical wiring or wood up there?
I hope a new run will fix it...but heat is a stretch.
Ordinarily I would agree, but I think even PVC insulators can start to suffer from thermal breakdown if exposed to 150* for extended periods of time. Given that cable has been in that attic - at least part of it for 20 years - replacing the entire run from the demarc to the drop isn't a meritless idea.
Putting a final bow on this issue:1. We did have confirmed upstream signal issues from Cox. Those have been fixed.
2. The tech ran a completely new cable from the demarc to my wall outlet.
3. Signal tests showed perfect for strength and throughput.
4. Connected from cablemodem to WIndows laptop - speeds perfect.
5. Moved connection from cablemodem back to firewall appliance - speeds throttled down to about 25% of normal.
6. Chatted with Cox tech, reviewed situation, sent someone out to double-check speeds and signal strengths. THey did, all good.
7. Reconnected to my firewall appliance, low speed returns.
8. Temporarily reconfigured firewall appliance to get direct connection from cablemodem rather than through virtual machine hosting firewall: full speeds realized
9. I was able to do some further internal performance testing and confirmed the problem has arisen somewhere within the configuration of my firewall appliance, which is decidedly a non-standard beast. Obviously not Cox's job to support that.
Bottom line, Cox is delivering what they promised, and i have to research the slowdown issue on my end.
Thanks to everyone who tried to help in this thread.
Let's see some perfect signals!
I am more curious what kind of "firewall appliance" was being troublesome. Firewalla?
I don't think it's a specific brand but perhaps the type of firewall. Firewalls can get quite granular.