Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: 1 2 3 [4] 5

Author Topic: Line 3 Upstream (Again)  (Read 9856 times)

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #45 on: April 08, 2021, 01:36:50 AM »

Someone ought to be able to work this out. Why only line 3 upstream? And in fact why is it only upstream that is incredibly slow? (Mind you, this is having not yet properly looked at any downstream issues since the drop cable swap-out, regarding packet loss.)

And, agrees with Burakkucat :)

Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Line 3 Upstream (Again)
« Reply #46 on: April 08, 2021, 04:17:12 PM »

In all seriousness, if senior management (and yourself) could cope without Internet access for, say, one hour it might be a worthwhile experiment to power off each modem and then disconnect each one from its respective NTE5. Once the "off time" has elapsed, power on each modem and then reconnect each one to its respective NTE5.
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #47 on: April 09, 2021, 02:20:17 AM »

Will do. I’d be interested to hear more about your thinking behind leaving a modem powered off for the duration vs leaving it powered up but merely disconnected.

I’ll recruit senior management for this task in the morning.
Logged

tubaman

  • Senior Kitizen
  • ******
  • Posts: 12507
Re: Line 3 Upstream (Again)
« Reply #48 on: April 09, 2021, 08:16:08 AM »

It is curious that the fault appeared gone after the drop-wire was replaced but returned a few hours later. This suggests to me either a poor joint/connection somewhere or a modem starting to fail once 'warmed up'.
If you've already tried swapping the modem and power supply, which I seem to remember you have, then I'd try disconnecting and reconnecting both ends of the line cord a few times to clean those connections and then another line cord if that made no difference. At least then you have proved that none of your equipment is the cause and that it is definitely an Openreach issue.
Logged
BT FTTC 55/10 Huawei Cab - Zyxel VMG8924-B10A

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #49 on: April 09, 2021, 01:06:50 PM »

I haven’t swapped over the modem or its psu recently. I just think that I spoke too soon the other day, saying that all was good; it was good/bad for various periods, that’s my my best guess.

AA has suggested that I disconnect the modem for a while so that it doesn’t muck up the overall bonded connection, saying that it’s pointless to try sending anything upstream up a 61k sync rate link. AA is going to book one final Openreach engineer on it and then we have agreed that we will give up if no success.
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Line 3 Upstream (Again)
« Reply #50 on: April 09, 2021, 11:18:31 PM »

Will do. I’d be interested to hear more about your thinking behind leaving a modem powered off for the duration . . .

The modem will then perform a cold boot, with all counters, etc, initialised to zero and the physical electronics in a (temperature) cold state. As tubaman has mentioned, above.
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #51 on: April 10, 2021, 11:36:59 AM »

Oh I see, I didn’t get the literal temperature thing. Good point, thanks Tubaman, have done as you suggest.
Logged

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #52 on: April 11, 2021, 06:14:18 AM »

The engineer from OR arrived on Saturday afternoon. Janet really needed my help in speaking the tech words to the engineer, so I wobbled out of my bed and was manoeuvred into the infrastructure- and servers room, aka Janet's dressing room. I explained very carefully to the engineer what the fault was, its importance and the background. He really ‘got it’; it was quite clear that he understood perfectly, but he didn’t have a solution. Fortunately there was something else wrong besides the agreed appalling u/s sync rate, which was a number of ES upstream during a 5 min long standard test. (Is it called a "close-out" test?) Because of this fail on ES and the general mysterious vexing nature of the problem, he phoned BTW support 2nd level or whatever it’s called; again couldn’t hear half of it; Anyway our man was escalated past the first-line representatives, but couldn’t get anything useful. He thought about changing target SNRM but I told him that that wouldn’t make a remotely big enough difference, which indeed turned out to be the case. He asked me about capping of the upstream - I had forgotten that there is such a thing for ADSL2 upstream - and I said that no one had applied any such thing, and it would show up in clueless’ logs anyway.

After we had discussed it for a little while, our engineer then left, saying he was heading to the exchange.

Whatever he then did, he somehow had fixed the problem and phoned back, so I gave him the good news that I agreed that it was looking superb at my end too! Even if he didn’t know how; he did something. Back up to 399 kbps @ 6dB. So a superb result.

Because the sound quality was better, yet totally appalling, the engineer's call to BTW was ‘audible’ to me (barely) on the phone’s loudspeaker. I asked the engineer about the audio quality of these calls and he said it was a continuing pain in the neck. At times we couldn’t make out anything at  all, due to a combination of bad call quality and very heavy Indian accents. It was very interesting to get an insight into what an engineer gets up to when he/she is in a call; I’ve never been in attendance before.

So line 3 has been saved from the chop.  ;D

I just wish there were fewer mysteries.
Logged

tubaman

  • Senior Kitizen
  • ******
  • Posts: 12507
Re: Line 3 Upstream (Again)
« Reply #53 on: April 11, 2021, 09:32:44 AM »

Let's hope it remains fixed as issues like this that go away but you're not sure why are really frustrating. I suspect he reseated a card and/or connection of some sort in the exchange.
 :)
Logged
BT FTTC 55/10 Huawei Cab - Zyxel VMG8924-B10A

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Line 3 Upstream (Again)
« Reply #54 on: April 11, 2021, 10:22:07 PM »

Hopefully that good news is not short-term.  :)

I wonder what feedback, regarding the engineering visit, A&A will receive?
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #55 on: April 12, 2021, 07:16:27 AM »

AA should get the engineer’s notes and they have already had my full write up, which unfortunately omits the crucial last chapter if there is any, if anything was known about any solution found.

I am know left with another problem for AA to look at, that of dripping blood on line 3.
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Line 3 Upstream (Again)
« Reply #56 on: May 20, 2021, 10:14:35 PM »

[Moderator note: This post and the five that follow it have been split off from the AA MSO topic.]

I get three times as many messages because I have three lines.

What has happened to number four?
« Last Edit: May 22, 2021, 09:31:36 PM by burakkucat »
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #57 on: May 22, 2021, 07:54:59 AM »

Line 3 upstream was so bad that I killed it. There was no way of getting OR too do the right thing as they don’t give any kind of upstream guarantee, which is ridiculous as they should treat it the same as downstream. I agreed with AA that we would get nowhere with OR after an unsuccessful call-out. Line 3 was always slow on upstream, with only ~400kbps as opposed to >500k sync rate. But other than that it was ok until a month or so ago. I also decided that there was a possibility that upstream might be affecting TCP and I mentioned this to AA who agreed that it was not an unreasonable thought. The sync rate had been ~150k @ 6dB or even worse at higher SNR. Downstream was excellent. I should really split this off as an obituary thread.

I’m thinking of getting a new line but guess what, I’m bound to get exactly the same pair that I’ve just given up, knowing my luck. I could of course just risk it, as what’s the worst that could happen. The pair I’ve given up would be fine for POTS usage, if anyone still does that. I outfit also need to organise the engineer a bit to make share that she or he doesn’t give me an unneeded drop-cable.
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Line 3 Upstream (Again)
« Reply #58 on: May 22, 2021, 05:30:28 PM »

Line 3 upstream was so bad that I killed it.

Looking in my notes, are you referring to the third line you had installed (i.e. drop cable 2, pair 1) or the line carrying the *@a.3 service (i.e. drop cable 1, pair 2)?
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Line 3 Upstream (Again)
« Reply #59 on: May 22, 2021, 06:22:58 PM »

It is the line carrying the *@a.3 service that is no more and ‘[I think that is indeed drop cable 1, pair 2, just as you said.

I had been making changes (upgrades in flexibility) to software that I have written for the iPad that allow for a non-dense, one-to-on mapping between so-called "slot ids" ie the number in ‘*@a.3" AA’s service identifier and the interface object that it maps to in the FB2900 and the associated VLANs that connect from the FB2900 to each modem. With no #3 any more, the VLAN numbers are 101, 102, 104 now (vlan=100+slot_id, or 100-1+slot_id, I forget which). Originally the slot id numbers were assumed to always be consecutive and that had to be changed so several new functions were written mapping a given slot id to one of various link/modem property values and these functions hid the details of such mappings. These get_property_xx( slot_id ) functions are driven by a small database. Even though these changes were made in readiness, some bugs were still found when a non-trivial mapping came into being with the deletion of the various slot-id #3 objects leaving 1, 2, 4. Mappings use the same slot ids as before, it’s just that there is a hole/gap in the set of input values to the function, as there is no longer any slot=3. Debugging really subtle bugs in the middle of the night. Ho hum.
Logged
Pages: 1 2 3 [4] 5