120M : 152M - Massive ammount of errors & No Upload STM & Internet Loss

[DanielMeah]

Anyone have an idea what is causing such a problem?

Downloads at full speed. but lots of Pre RS Errors on channel 50. on top of that upload STM seems to be disabled. This error seems to have also caused random internet loss every now and then for around 20 seconds (complete loss) but the connection stays locked (i.e the sync and uptime)



Upload seems fine too



Uptime


Operational Config:

[General Maximus]

pre-rs errors don't matter so most, it is the port errors you need to worry about. Have you tried rebooting the shub to see if it locks on to different channels? Your power levels are fine, I haven't seen a 6 channel bonding group in a while, I thought that everyone was on 8 with all the speed upgrades that have been going on.

[DanielMeah]

I got upgraded via retentions (My area was 152Mb ready but for some reason upgrade for existing wasnt till june 2015)

i've tried several times rebooting the hub (support couldnt fix problem)

other day i had nearly 80k errors in less than 10 mins!

[General Maximus]

you can try ringing tech support and see if they will book a tech out but afaik the rs errors are normally a result of tired lasers so the tech would only come out and escalate it.

[jb66]

The tech cant escalate it unless there is a fault

[General Maximus]

is there a tool or some piece of equipment that measures the lasers output/strength/whatever you measure on a laser that you use?

[jb66]

Originally Posted by General Maximus

is there a tool or some piece of equipment that measures the lasers output/strength/whatever you measure on a laser that you use?

If it is an area issue second line would have to identify and escalate it to networks, when a field tech arrives he only has a simple signal meter which can only measure downstream power levels. Unless it's blatantly obvious he can't escalate it as he can't prove there is a netwprk fault.

[General Maximus]

how do you go about getting like this resolved then? Every time I have heard rs error mentioned Seph has always said tired lasers or amps on their way out. Surely there is only a finite amount of equipment that can be a contributing factor to this sort of thing and if you have no way of directly testing it and you can't escalate it how does it get fixed? (I am not being sarcastic, it is a genuine question out of curiosity)

[jb66]

Originally Posted by General Maximus

how do you go about getting like this resolved then? Every time I have heard rs error mentioned Seph has always said tired lasers or amps on their way out. Surely there is only a finite amount of equipment that can be a contributing factor to this sort of thing and if you have no way of directly testing it and you can't escalate it how does it get fixed? (I am not being sarcastic, it is a genuine question out of curiosity)

It's a very good question.

We can see what the upstream power level is and snr by using the customers equipment. We have no way of seeing t3 timeouts, power adjustments, insertion fails, pre/post rs errors.

We trust that second line compares the customers connection with the others on the same ubr. If multiple modems all have issues then they raise it to networks. If it's just one customer then we check connectors are tight and not corroded. Also check signal loss and swap the hub if needed

[General Maximus]

interesting, I can see it going round in circles and taking a while if you have to rely on 2nd line to diagnose it and then pass it on to networks and wait for them to look at it.

[horseman]

Originally Posted by DanielMeah

Anyone have an idea what is causing such a problem?....

Yes several but speculation is premature without the further missing diagnostics:

1. The corresponding Network Log
2. System uptime from Advanced Settings > Device Status (is it the same as Session Time?)
3. When you perceive a disco then can you still ping (say 8.8.8.8) from SHub? (Advanced Settings > Tools)

The pre-RS errors are hardware corrected in-flight but serve as a "warning" in the appropriate context. The post-RS( which would potentially represent TCP retransmits or lost UDP packets due to uncorrectable FEC) depends on the rate of increase and should be (hopefully) coupled with any T4(or other error) stanza's in network log.

[DanielMeah]

1, is almost empty, nothing of importance.
2. everything stays connected and same uptime.
3. when this happens nothing works. pings stop etc.

as of now channel 50 is nearly in 500k

[General Maximus]

I have seen over 1 million before

Edit: this is the thread I was thinking of http://www.cableforum.co.uk/board/12...rs-errors.html

[sollp]

Originally Posted by General Maximus

is there a tool or some piece of equipment that measures the lasers output/strength/whatever you measure on a laser that you use?

Once the fault has been referred to networks, they can check all this and adjust replace as necessary. Unfortunately the Service Tech's only have a very basic test meter that will only tell them the Downstream levels nothing else.

They must be able to convince at times the network Tech he is referring the fault to that there is an issue.

[DanielMeah]

Issue Resolved, apparently it was a "Local" network problem. (the engineer told me it was further in the network rather than the cabinet)

now youtube is failing lol.