Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[tiffy]

Unfortunately, as of approx. 01:00 hrs. this morning the US SNRM oscillations have returned, line did not drop/re-synch.
US ES's & SES's are on the rise associated with the disturbance, U1 bit swaps very high etc..
DSLStats graphs attached for reference.

As still no response from Plusnet to my ticket raised on 8/4/20 contacted tech. support by phone.
Surprisingly short wait, tech. support rep. was reasonably well informed and after viewing the DSLStats fault graphs I had attached to the ticket agreed there was an issue, however nothing showed up on his available line tests other than a decrease in US synch. speed.

Agreed to pass the issue on to higher level tech. support PN/BT for further investigation.
Hopefully the fault won't disappear again now pending investigation assuming that anything will actually show up by external testing.

[tiffy]

Received feedback from Plusnet, higher level tech. support department via phone call.

Tech. did agree that there certainly did appear to be an issue, probably intermittant, even though nothing was showing up in any of their available tests, stated that under "normal" circumstances this would be refered to BT/OR for further testing, however, under current restrictions this was not possible as my service was still "fit for purpose" and did not fall into the urgent catagory under current rules.

Asked if a DLM reset was possible but was advised that they can no longer submit requests to BT/OR for this.

So, looks like I'am stuck with the situation until restrictions are relaxed, can just hope my US remains usable and DS remains unaffected.
Can't really fault PN as they have promptly followed up my query (after making a phone call) and they are working within the restrictions imposed at present by something outwith anyones control.

[Chrysalis]

if dlm resets are now been refused i feel interleaving profiles either need to be disabled or the old dlm put back with agressive recovery times

thanks for keeping us updated

aaisp told me the same regarding resets

[tiffy]

Yes, it's all a bit worrying.

The PN tech. support agent I spoke to was very apologetic regarding my situation but stated that all staff were instructed to prioritise all issues reported and really only action those deemed "high priority" such as vulnerable, isolated people and perhaps someone forced to work from home if this could be proven.
In theory, we are all isolated at present and vulnerability is very much open to interpretation !

I did ask if my fault did progress to a disconnection and loss of ability to use my BB would that be addressed ?
Didn't really get a straight answer on that as he stated that any requirement by BT/OR to enter my residence for fault diagnosis would not be possible at present.
This is when I requested a DLM reset which was refused.

Although my fault is currently not impacting usage it is certainly getting worse, see attached graphs, can now see that the increasing US SNRM oscillations are beginning to impact DS SNRM, US ES's & SES's are increasing steadily.

In a way, it might be best if the system "broke" completely, however, would I then be in the catagory to merit remedial action ?

Edit: Typo corrected.

[Chrysalis]

A DLM reset should be able to be action'd remotely, so in theory from someone working from home, I can understand engineer staffing restrictions as those are of risk of infection, but DLM resets been suspended shouldnt be required if openreach organised their staff properly.

I expect if your line went down, they would try to fix remotely, and if that wasnt enough then the vulnerability stuff might come into it.

[tiffy]

A DLM reset should be able to be action'd remotely, so in theory from someone working from home

That is my, somewhat limited, understanding also, DLM resets appeared to be relatively easy to obtain from Plusnet last year judging by posts on their forum, not the case any more apparently.

In my case and under the current BT/OR attendance restrictions, thought that a DLM reset was something that could be tried on my line as nothing was showing up on PN's testing and would avoid any initial requirement for attendance.

Over night, have had a DLM re-synch @ 01:00 hrs, US G.Inp has been activated and I've obviously lost more US synch speed, now down to 1.7 Mbps from original, pre-fault speed of > 6 Mbps..
Still maintaining reasonable DS synch speed with Re-Tx high still active, DS SNRM still showing minor oscillations which appear to be in sympathy with the much greater US SNRM amplitude oscillations. 

[Chrysalis]

in your case i dont think a dlm reset might offer much though, there seems to be some kind of underlying issue

[burakkucat]

I am beginning to wonder if the line-card port has "gone bad". Normally an overnight "relax" would clear such oscillations.

[tiffy]

in your case i dont think a dlm reset might offer much though, there seems to be some kind of underlying issue

Yes, you are probably correct, although PN conceded from the DSLStats graphs supplied that there did appear to be an issue were adamant that they could not action any further at this time as a BT engineer would require home access for testing purposes, this is why I requested a DLM reset as something which could be performed remotely and would be unlikely to make my situation any worse if it did not effect a cure.
PN also insisted on closing out the fault ticket as "resolved" although I protested strongly that this was not the case, advised that I should re-submit the fault if necessary when isolation conditions improve.

I am beginning to wonder if the line-card port has "gone bad". Normally an overnight "relax" would clear such oscillations.

Quite possible I'am sure, have implemented two "overnight relaxes" on the line since the issue appeared to no avail.
The fault certainly appears to have an intermittent element as can be observed on the last 24 hr. SNRM graph, attached, US SNRM oscillations decrease in amplitude then just went away early this morning without any further re-synch..
This pattern occured a few days ago then fault just returned again with a vengance ?

Could likely now improve my US synch speed with a re-synch under the current stable conditions, however, will resist the temptation and let things settle as I'am sure the oscillations will return again.
G.Inp (Re-Tx) remains applied US as does DS with high Re-Tx, both ES's practically zero.

I have a spare ZyXEL VMG1312-B10D configured ready for use which I can try on the line but really don't think it will do any good and certainly won't perform as well as my current 1312-B10A, really don't believe that I have a hardware issue at this end.
The Billion modem/router I substituted a few days ago appeared not to like my line and produced a major amount of DS errors unrelated to the current US issue, this unit subsequently is working perfectly on my daughters VDSL 40/10 line and out performs her ZyXEL 1312D ?   

[tiffy]

The cycle continues.

Had a DLM re-synch @ 02:30 this morning, I presume this was to remove US G.Inp as US error rate was practically zero since G.Inp application and the SNRM oscillations abating, know that historically DLM action removes US G.Inp quite quickly once error rates lower.

Have gained a bit more US synch speed but US SNRM oscillations have returned on DLM re-synch, US error rate is now building up again without G.Inp and the current unstable SNRM so probably heading for another G.Inp application/removal cycle somewhere down the line !

As can be observed on previous posts, sometimes the fault appears/disappears by itself not attributable to any visable event.
Sometimes re-synching triggers re-appearance of the fault sometimes does not.
The only thing I can see that remains consistant througout is the change in QLN US profile from before to after the issue.

Attached, todays SNRM & QLN graphs.
Historical QLN graph pre-issue.   

[tiffy]

Apologies, attached HLog pre-issue graph in previous post, should have been QLN pre-issue graph, now attached.

[Chrysalis]

Oscillations are bad, I was looking at my line yesterday and determined could correlate bitswapping activity with errors, e.g. when I have higher bitswaps/min the FEC is higher.

After making that link, I checked the bitswaps/tone graph and noticed a few tones had abnormal levels of bitswapping, then checked the SNR per tone graph and noticed a oscillation I hadnt noticed before as it doesnt stand out that much.

So when a line has large levels of oscillation I think thats an underlying issue.

[burakkucat]

So when a line has large levels of oscillation I think thats an underlying issue.

Yes, agreed. The ongoing mystery is: "What is the cause?" 

[rjpreston]

i have same / v similar issue (rapid and semi-continuous instability of Up SNR leading to resyncs, massive reduction of Up speed (normally 8Mb reduced to 0.9Mb); and corresponding but less severe reduction of Down speed (33Mb down to 27Mb; using a Zyxel VMG1312-B10A). The charts that the OP has posted already could have been mine they are so similar! This has been going on since mid-Feb and is steadily getting worse. I even got a 2nd modem from ebay just in case the first was faulty, but the condition remains.

I logged a fault with my ISP (BT) expecting to have to wait months until lockdown is over. But they sent a technician out yesterday: A very weird experience of having someone in your home at the moment! A fault on the line was identified and he made another appointment for 1-6pm today to fix it. Watch this space...

[tiffy]

@rjpreston

Many thanks for your input, very interesting.

All I will say at present is count yourself lucky your ISP is BT, I am just about to post what I believe is the resolution of my line issus and this will explain my meaning.