cancel
Showing results for 
Search instead for 
Did you mean: 

How common are post rs errors?

alphaomega16
Up to speed

If I CTRL+F5 for refresh and load router status for the first time im seeing post error counts alongside a lot of pre.

alphaomega16_0-1671048111235.png

 

alphaomega16_0-1671048181730.png

alphaomega16_0-1671048251800.png

 

Every time I check it seems to get worse

alphaomega16_0-1671048469196.png

 

Edit

Upstream seems to have gone from 16 to 32 now and this website loads a lot faster now.

alphaomega16_0-1671049520915.png

My cable is still in a green pipe above ground, could cold be causing issues?  Wouldn't have thought so as I got a lan cable going on the outside wall to downstairs PC which is fine.

 

Edit

Just done another check, all upstreams are at 64 now but

alphaomega16_0-1671050474694.png

Never seen post errors before

Another edit

upstream 5 back to 16qam

62 REPLIES 62

It is perversely amusing to see the Forum Team sort of blaming the Community for the time it takes for VM to do their job properly. For a start, their stupid algorithms are not properly taking age of problem into account and clearly no AI is applied to content to assist with the priority of an ongoing issue.

That area noise problem has been present from day one (imo) and nothing’s been done about it. The SNR paragraphs in Molly’s reply are the true items of value, but probably with undue attribution of 3rd party equipment attached to the network. It’s very rarely that and it would be helpful if they could take the trouble to warn us of particular circumstances of past occurrence that should be avoided. Routers and switches attached to a network don’t create noise; mains spikes would cause uniform glitches and that would be across the piste in any locality. El cheapo fridge thermostats have been know to cause periodic noise ingress that could glitch right across a locality for a few moments every hour. In that sense, “customer equipment” is correct.

As to tracking noise down, as I’ve explained before (and VM have given no explanation), the CMTS is pre-configured to tolerate noise on the upstream up to a threshold that sets 16QAM on the affected channels. It’s what you don’t see in the stats that causes the resets - when it falls below 16QAM. The fact that it has occurred on several channels (occasionally) at different frequencies give VM a headache. Thus taking a long time to chase down and fix is absolutely accurate.

What fails here is any kind of worthwhile update as to progress. You don’t even know if they are working on it. For this case, I suggest you request a Fault Ticket reference number. It’s the only way you’ll know that it’s being worked on.

Seph - ( DEFROCKED - My advice is at your risk)

5th channel I noticed at 32QAM a couple of times yesterday but has gone back to 16

alphaomega16_0-1672490629529.png

 

Not getting as many timeouts currently but still seems quite noisy on line.

Sat Dec 31 12:21:28 20223No Ranging Response received - T3 time-out;CM-MAC=**:**:**:**:**:**;CMTS-MAC=**:**:**:**:**:**;CM-QOS=1.1;CM-VER=3.1;

Ask for the Ticket Reference. They’ll ignore me. 

 

Seph - ( DEFROCKED - My advice is at your risk)

Forgot to check yesterday but apparently Channel 3 is in the lead with timeouts this time around and its a miracle #5 has moved from 16qam

alphaomega16_0-1672636573676.png

Sat Dec 31 12:21:28 20223No Ranging Response received - T3 time-out;CM-MAC=**:**:**:**:**:**;CMTS-MAC=**:**:**:**:**:**;CM-QOS=1.1;CM-VER=3.1;

 

Edit

Ok, thats odd, just noticed that it says 31st still.

I like to give as much info as possible and you can see I have posted image of each 3 I have gotten and tried to do the time I got them as well but it seems a date is missing in the log?

Unsure of when it occurred.

The odd T3 or two are neither here nor there.  We all get them at some time or another. They are low priority events (being keep-Alice’s - somewhat akin to pings) and kickers get twisted only when the 16 consecutive have occurred.

But the hopping around of upstream QAM is symptomatic up a chronic upstream noise situation, as you will have gathered.  Normally, VM are onto this (noise) but they won’t proactively tell you.  Nor have the Forum Team, who have the mans of finding out if there’s a fault ticket raised, said anything worthwhile. 

Technically, the network is behaving correctly in that the symbol density (QAM) is reduced by the CMTS when there is noise.  But the circuit breaks occur when the 16QAM tolerance is breached.  VM’s logs in their CMTS will have all that information.  

But’s what’s so really bad is the failure of the so-called support team on this forum to competently assist in resolution.  I suppose we can excuse the individuals - they have to follow policy.  But the masters who thought that the Forums would be a decent support medium need to re-evaluate.

Anyway, I’ve said all that before and I remain defrocked!

 

Seph - ( DEFROCKED - My advice is at your risk)

Understood Seph, ill hold off any updates unless anything spikes out of the ordinary or gets worse but I hope whatever is going on with the upstream is being looked into.

Unless they give you a ticket reference, you have no way of knowing if they're onto it. As explained before, though, line noise is very tricky to pin down if it's not generated at the CMTS end.

If your circuit appears to have dropped, the BQM would show a tall red spike. If it doesn't, then the Hub could be faulty in the non-modem section.
Seph - ( DEFROCKED - My advice is at your risk)


@Sephiroth wrote:

Snip...

I suppose we can excuse the individuals - they have to follow policy.  But the masters who thought that the Forums would be a decent support medium need to re-evaluate.

...Snip

 


No surely not....!😉

On another matter, the 23.6 MHz upstream channel does seem to be problematic regarding noise impairment, I thought it was going to be phased out, is that not the case, or is the process just taking a bit longer than expected?

And to everyone else reading this thread, as I'm sure Seph already knows and can probably skip the rest of this ....

To be honest here, VM are really only concerned about what the contracts require them to provide, and if you read the small print, that's not too much in effect!

Basically are you getting better than 50% of the rated download speed averaged over a period of three days? If 'yes' then 'just go away, you annoying person - although, your custom is important to us and we do appreciate your monthly subscriptions!''.

Basically VM make no promises about upload speed, latency or jitter (gamers, take note here)!

The underlaying technology involved here is called DOCSIS and it does get a bit of a bad press sometimes, but it really is quite robust and designed to try to keep working and maintain a connection, whatever. The problems can be when people look at the logs or look at the stats and assume that it has all gone 'Pete Tong' when actually it's just the system reconfiguring itself to cope with changing conditions.

And you know what, that's really all fine - nobody really should expect that every customer should understand how the underlying technology works or what these (sometimes) scary looking messages really mean. And that's where we come in, because some of us, actually do understand all this stuff (arguably better than, oh, say 90% of every VM employee and certainly more than every single member of the forum team here); what it means, and, more importantly, are able to translate it into terms that non-technical people can get and either reassure them not to worry, or alternatively, say, well that's really not too good, VM's customer services are jerking your chain, and you probably need to do xyz.....

Basically, DOCSIS tries to sort itself out, channels dropping to 32 or 16 QAM absolutely 'are' indicative of noise ingress or otherwise impairment - but does that necessarily reduce your connection quality? If no, then VM aren't going to pull the stops out to fix it, but it is worth keeping an eye on to see if it gets worse, to the point when it does impair your connection and that's when you can start kicking off and escalate to the industry adjudicator. Evidence that this has been going on for some time and VM hasn't bothered to address it, would mitigate in your favour! 

 

Whilst the last upstream channel is still at 16qam it seems to have been fine until tonight I noticed my latency has jumped up in games and when checking im seeing this

 

alphaomega16_0-1673915080389.png

This is usually when I start getting grief with my connection, anybody have any idea why this happens?

Hopefully this isn't the start of more issues.

Edit

Didn't last for hours this time, was short bump but this is whilst my connection is pretty much idle.

alphaomega16_0-1673916192545.png

 

These things happen when you’re not looking. Either noise exceeds the tolerance threshold or there is contention on that set of bonded channels. Indeed, contention increases when a channel becomes noisy, particularly on the upstream. 

 

Seph - ( DEFROCKED - My advice is at your risk)