I've been tracking an issue while we've had some unusually warm weather in this area and comcast business modem & voice resetting simultaneously.
(This isn't really a truck roll moment, it's more of a "hey are you guys aware of this" thing)
If the temperature in this area (zip 98055) goes above 80F, there seem to be periodic resyncs/resets with the connections. (I would hazard a guess that some of the equipment is in a non-temperature controlled room and eventually just starts wigging out).
A few things seem to support this theory:
1. I've been tracking outside temperature along with these (usually) brief outages
2. They seem to begin in the late afternoon (around 3pm ~ 4pm local time) when the temperature is near it's peak.
3. They continue, semi-randomly, until around midnight when it's really starting to cool down.
1. 2014-07-01 16:57 outside temp: 93F
2. 2014-07-01 20:55 outside temp: 83F
3. 2014-07-02 15:31 outside temp: 81F
4. 2014-07-02 15:55 outside temp: 81F
5. 2014-07-02 16:03 outside temp: 82F
6. 2014-07-02 20:03 outside temp: 76F
7. 2014-07-06 15:50 outside temp: 87F (this was an especially ugly one, didn't get sync back for 23 minutes)
Anyway you get the idea.
My home office here is air conditioned but air conditioning isn't very common in this area so I can totally see that maybe somewhere up the line something is overheating.
There's also a lot of construction in this area so there could be some issues related to that too.
I believe the equipment and connection at the home office here is fine (and it's a 2012 install) but just have been noticing some possibly temperature related issues.
Solved! Go to Solution.
Temperature is something that generally only affects client side equipment as all server equipment is hosted in data centers which are kept extremely cold.
Just a wild guess here, but it is possible that the heat is causing a specific section of cable to expand and compress which is causing some issues. Generally only damaged cable would have this issue, though, and I'd expect Comcast to be able to se that as it'd be putting a lot of noise on the line.
kraze is right about the datacenter equipment, and it sounds like you're keeping your end's equipment cool enough. Think you could post your modem signal levels? I believe that higher temperatures cause upstream power levels to decrease and downstream power to increase, and they may be borderline in normal weather but drifting out of spec/out of ideal range when it's hot outside. You didn't mention your modem model, but you can find these signals generally by browsing to http://10.1.10.1, http://10.0.0.1, or http://192.168.100.1, or whichever IP address your modem has.
Modem: SMC SMCD3G-CCR (Firmware 18.104.22.168)
Stats as of right now (connection is fine):
Outside Temp: 63F
Downstream Power: -4.363595 dBmV, -4.589916 dBmV, -4.245671 dBmV, -5.448189 dBmV
Upstream Power: 47.0000 dBmV, 48.5000 dBmV, 49.0000 dBmV, 48.5000 dBmV
Downstream SNR: 37.355988 dB, 36.844463 dB, 37.355988 dB, 37.092701 dB
downstream power: -5.396182 dBmV, -5.507679 dBmV, -5.218738 dBmV, -6.382004 dBmV
downstream SNR: 37.355988 dB, 36.844463 dB, 37.092701 dB, 36.844463 dB
upstream power: 47.0000 dBmV, 48.5000 dBmV, 49.0000 dBmV, 48.5000 dBmV
Had the first outage issue of the day, but the numbers seem ok (?):
Outside temp at the time of the outage was around 86F.
Unfortunately I wasn't able to jump on the cablemodem during it but I did catch it 15 minutes after it stopped.
2014-07-08 15:22 (didn't route until 15:26)
2014-07-08 15:30 (didn't route until 15:33)
2014-07-08 15:37 (didn't route until 15:41)
2014-07-08 15:44 (didn't route until 15:45)
It recovered, eventualy, on it's own, usually when it does this "4 minute" pattern I have to power cycle it (off for 2m).
Modem Stats (shortly after event):
2014-07-08 16:00 modem: downstream power: -5.127232 dBmV, -5.203783 dBmV, -5.140637 dBmV , -6.127459 dBmV
2014-07-08 16:00 modem: downstream SNR: 37.355988 dB, 36.844463 dB, 37.092701 dB, 36.844463 dB
2014-07-08 16:00 modem: upstream power: 47.0000 dBmV, 48.5000 dBmV, 48.5000 dBmV, 48.5000 dBmV
Modem Stats Before (I knew I should've done a sample at 15:00):
2014-07-08 14:00 modem: downstream power: -5.147938 dBmV, -5.242950 dBmV, -5.159728 dBmV, -6.423411 dBmV
2014-07-08 14:00 modem: downstream SNR: 37.355988 dB, 36.844463 dB, 37.092701 dB, 36.844463 dB
2014-07-08 14:00 modem: upstream power: 47.0000 dBmV, 48.5000 dBmV, 49.0000 dBmV, 48.5000 dBmV
(Again this isn't really a truck roll thing as it doesn't do it all the time, seems to be more frequent when it's really warm outside, like today. I'm mostly just curious what the heck is going on with it.)
What's the optimal baseline for power and SNR?
Just had another one and this time, I got better info:
2014-07-07 16:23 start: alerter (couldn't ping gateway) (outside temp: 86F)
2014-07-08 16:23 modem: downstream power: -5.517918 dBmV, -5.546569 dBmV, -5.458819 dBmV, -6.550837 dBmV
2014-07-08 16:23 modem: downstream SNR: 37.355988 dB, 36.844463 dB, 37.092701 dB, 36.844463 dB
2014-07-08 16:23 modem: upstream power: 47.0000 dBmV (only channel 3)
2014-07-08 16:30 environment: 86F
2014-07-07 16:39 event: power cycle (unplug)
2014-07-07 16:42 event: power on
2014-07-07 16:44 event: connected
2014-07-07 16:45 end: online
2014-07-08 16:45 modem: downstream power: -5.070086 dBmV, -5.203783 dBmV, -4.964192 dBmV, -6.232115 dBmV
2014-07-08 16:45 modem: downstream SNR: 37.355988 dB, 36.844463 dB, 37.092701 dB, 36.844463 dB
2014-07-08 16:45 modem: upstream power: 47.5000 dBmV, 49.0000 dBmV, 49.0000 dBmV, 47.5000 dBmV
Downstream Frequencies: 597.001831 MHz, 578.998047 MHz, 584.999268 MHz, 591.000610 MHz
Upstream Frequencies: 18900000 Hz (Ch 3), 30100000 Hz (Ch 1), 36500000 Hz (Ch 2), 23700000 Hz (Ch 4)
During the outage I pinged the modem's static WAN gateway (22.214.171.124) and it was responding. So it appears the outage is the loss of transmit ability. (This particular one anyway).
(Note the delay with power cycling etc was so I could poke around the modem admin and write stuff down. I'm normally much more responsive than that )
During the outage the first three modem lights (other than power) from left to right were all steady on, it basically thought it was connected but wasn't transmitting until I power cycled it.
Your upstream signal seems to be too high. At 50 you could start seeing issues and it's possible it's dipping out of spec at certain parts of the day. Have you had a tech out at all?
(Edit: Why are Rich Text & HTML edit here when they don't work correctly?)
The thing I don't get, is we have comcast business in other locations (granted the modems aren't the same) and the "correct range" downstream/upstream power seems to vary.
Site A: (Netgear 1.04)
|Downstream Frequency||555000000 Hz||561000000 Hz||567000000 Hz||573000000 Hz||579000000 Hz||585000000 Hz||591000000 Hz||597000000 Hz|
|Modulation||256 QAM||256 QAM||256 QAM||256 QAM||256 QAM||256 QAM||256 QAM||256 QAM|
|Symbol Rate||5.360537 Msym/sec||5.360537 Msym/sec||5.360537 Msym/sec||5.360537 Msym/sec||5.360537 Msym/sec||5.360537 Msym/sec||5.360537 Msym/sec||5.360537 Msym/sec|
|Downstream Power||-3.7 dBmV||-3.7 dBmV||-3.7 dBmV||-3.9 dBmV||-3.9 dBmV||-3.6 dBmV||-3.7 dBmV||-3.8 dBmV|
|SNR||40.3 dBmV||40.3 dBmV||40.1 dBmV||39.9 dBmV||39.9 dBmV||40.1 dBmV||39.9 dBmV||39.9 dBmV|
|Upstream Frequency||30100000 Hz||36500000 Hz||18900000 Hz||23700000 Hz|
|Symbol Rate||5120 sym/sec||5120 sym/sec||2560 sym/sec||5120 sym/sec|
|Upstream Power||51.0 dBmV||51.0 dBmV||51.0 dBmV||51.0 dBmV|
The upstream is over 50 dBmv and oddly I've never seen it change from 51. No problems either.
I do notice the modulation is different.
|Vendor Name||SMC Networks|
|Downstream Frequency||639000000 Hz|
|Symbol Rate||5.360537 Msym/sec|
|Downstream Power||-7.3 dBmV|
|Upstream Frequency||30100000 Hz|
|Symbol Rate||5120000 sym/sec|
|Upstream Power||47.1 dBmV|
So for me, it seems fairly inconclusive. I've been checking my values here much more often and I do believe direct sunlight with higher temperatures (85+) is having some kind of effect but it's not consistent at all.
(I had it flip here when a minute before all 4 upstream power's were under 50 dBmv.)
And again, this is primarily an intellectual exercise, it isn't annoying enough for a truck roll or anything, it's just weird.
What is annoying this new issue: