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 ... 3 4 [5] 6 7 ... 11

Author Topic: My line might have issues?  (Read 39678 times)

NewtronStar

  • Kitizen
  • ****
  • Posts: 4898
Re: My line might have issues?
« Reply #60 on: October 19, 2017, 09:47:26 PM »

Just looking into this situation and says to myself could my line at 1 kilometer be improved by switching to an ISP like A&A, this line would get over 2000+ errored seconds on Fastpath if it was not for G.INP just wondering how an ISP could cure this with the help of Openreach ?
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #61 on: October 19, 2017, 10:44:03 PM »

Just looking into this situation and says to myself could my line at 1 kilometer be improved by switching to an ISP like A&A, this line would get over 2000+ errored seconds on Fastpath if it was not for G.INP just wondering how an ISP could cure this with the help of Openreach ?

Hmm, I guess it depends what AAISP would see regarding the performance of your line. That's quite a lot of error seconds.

In my fault notes they mention 'policy discards', which I assume is referring to the "Downstream Policing Discard Rate" mentioned on a GEA test. Perhaps it means something different though. It's a gamble switching though, but certainly for my situation I feel I've definitely made the right decision now. I thought maybe I'd get some kind of response like "DLM is doing its job by enabling error correction and you have a sync rate that's higher than BT's estimate", but no :). I'm not sure when I'll get an engineer visit, could be a surprise visit too for all I know (has happened once early in the morning, that wasn't too fun when all the phones suddenly started ringing none stop). I know my line's on service/care level 2.5 which apparently gives me higher priority and clear by end of next working day (Monday to Saturday inclusive). However that might only apply to PSTN faults, not FTTC.

WLR3Test WLR3_CIDT_Test: Pass Line Test OK. Dial tone OK ServiceLevel:2.5, MainFaultLocation:OK, FaultReportAdvised:N, AppointmentRequired:N, LineStability:, NetworkStability:, StabilityStatement:

I wish I had G.INP, but would G.INP mask the problem I'm having or would it be noticeable on AAISP's CQM or some other diagnostic. I'm not sure. Not having G.INP is probably a blessing in disguise for my current problem. I queried a concern about possible charges if broadband engineer visits and for some reason can't find a fault, but I was told that charges won't be a problem as they won't be paying anything extra to fix a fault with the service. All good news so far.
« Last Edit: October 19, 2017, 10:51:05 PM by Ixel »
Logged

NewtronStar

  • Kitizen
  • ****
  • Posts: 4898
Re: My line might have issues?
« Reply #62 on: October 19, 2017, 11:15:23 PM »

Not having G.INP is probably a blessing

The whole idea of G.INP was to try and eliminate those errored seconds and it does a very good but that is not an option for you.

Still think back to the old days when Kitz, Baldeagle1 and BC said to let your circuit become service affected and become a very noticeable fault on OR's JDSU or you will get a no fault found, now I don't know how A&A are able to persuade the OR engineer there is a fault when his/her JDSU says there is no fault who then becomes the engineer A&A or Openreach  ::)
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #63 on: October 20, 2017, 09:56:14 AM »

Broadband engineer has been, unfortunately 'no fault found'.

He mentioned his tester's modem is a bit inferior to a modern modem and so could sync slower, which then makes me wonder if his tester is not as sensitive to errors if it's syncing slower. He was also of the attitude that my line could do with interleaving and not to worry about the higher ping/lower speed caused by it. I had to tell him to do a DLM reset however, he was doing tests initially with interleaving still on! How is that going to accomplish anything haha.

I'm now waiting on AAISP's response to the situation. Currently on the locked ECI /r for now.

EDIT: Plugged in DrayTek Vigor 2860Vac, so far stats seem fair. Maybe the fault has disappeared magically?

https://pastebin.com/raw/fHTXgpf6

Only thing I've observed is that before I'm certain I had an 'R' (RFEC) value of 0 instead of 16 on the downstream. Maybe now that it's 16 it's doing a job of helping handle the errors that I was originally getting on fastpath with a downstream 'R' of 0? I notice I'm getting a reasonable quantity of FEC errors at the moment.
« Last Edit: October 20, 2017, 11:38:03 AM by Ixel »
Logged

WWWombat

  • Kitizen
  • ****
  • Posts: 1674
Re: My line might have issues?
« Reply #64 on: October 20, 2017, 11:53:35 AM »

EDIT: Plugged in DrayTek Vigor 2860Vac, so far stats seem fair. Maybe the fault has disappeared magically?

https://pastebin.com/raw/fHTXgpf6

Only thing I've observed is that before I'm certain I had an 'R' (RFEC) value of 0 instead of 16 on the downstream. Maybe now that it's 16 it's doing a job of helping handle the errors that I was originally getting on fastpath with a downstream 'R' of 0? I notice I'm getting a reasonable quantity of FEC errors at the moment.

I'm not certain about that format of output, so this might be wrong, but...

On broadcom stats, in the framing parameters section, R is the number of bytes being used for the FEC overhead and N is the RS block size (ie the total size of the block that FEC applies to). On my current setup with G.INP active, R=8 and N=139. 8 bytes out of every 139 are the FEC redundancy overhead.

If they represent the same thing on your line, you could well have the FEC process active, using 16 bytes out of 255. This would explain the non-zero FEC counter and FECS counter.

On broadcom stats, the closely-related interleaving statistics can be seen (as D and I) in the same section. For yours, I think I is "INTLVBLOCK", while D is "DS Interleave Depth" in the top section instead. You don't appear to be interleaved. It also looks like DLM has not asked for interleaving (INP=0 and INTLVDelay=0).

So you look to have interleaving off, but FEC turned on. Not often seen downstream, but very common upstream.
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #65 on: October 20, 2017, 12:04:09 PM »

I'm not certain about that format of output, so this might be wrong, but...

On broadcom stats, in the framing parameters section, R is the number of bytes being used for the FEC overhead and N is the RS block size (ie the total size of the block that FEC applies to). On my current setup with G.INP active, R=8 and N=139. 8 bytes out of every 139 are the FEC redundancy overhead.

If they represent the same thing on your line, you could well have the FEC process active, using 16 bytes out of 255. This would explain the non-zero FEC counter and FECS counter.

On broadcom stats, the closely-related interleaving statistics can be seen (as D and I) in the same section. For yours, I think I is "INTLVBLOCK", while D is "DS Interleave Depth" in the top section instead. You don't appear to be interleaved. It also looks like DLM has not asked for interleaving (INP=0 and INTLVDelay=0).

So you look to have interleaving off, but FEC turned on. Not often seen downstream, but very common upstream.

I see, interesting. Yeah I thought usually downstream doesn't have it hence normally no FEC errors when on fastpath. I've always seen it on upstream. Perhaps that's why I'm no longer seeing the amount of CRC errors/error seconds I was getting on the downstream. I will try the unlocked ECI /r with LEDE later on as well. I doubt I'll try the HG612 though as I feel it's not as compatible with the ECI cab.
Logged

Chrysalis

  • Content Team
  • Addicted Kitizen
  • *
  • Posts: 7382
  • VM Gig1 - AAISP L2TP
Re: My line might have issues?
« Reply #66 on: October 20, 2017, 05:38:01 PM »

Ahh so thats why you avoiding broadcom, as I was going to ask.

Everyone here is in a better position to help if you active on MDWS.

The problem you will have with openreach is they consider interleaving to be a "solution" not a "workaround/mask".  So unless interleaving pulls a line below the handback threshold they will consider it all good, so now you going to be reliant on AAISP fighting your case to tell openreach thats not acceptable.

In my opinion the error rate you previously reported is way excessive, getting 1000s even in a single hour, however if the line is stable when interleaved, from a dumbed down view point openreach may just say its working as described and then refuse to do anything.  So aaisp will need to keep pushing them probably.

Did you tell aaisp about what you was told by your installation engineer?
« Last Edit: October 20, 2017, 05:40:05 PM by Chrysalis »
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #67 on: October 20, 2017, 05:55:09 PM »

Ahh so thats why you avoiding broadcom, as I was going to ask.

Everyone here is in a better position to help if you active on MDWS.

The problem you will have with openreach is they consider interleaving to be a "solution" not a "workaround/mask".  So unless interleaving pulls a line below the handback threshold they will consider it all good, so now you going to be reliant on AAISP fighting your case to tell openreach thats not acceptable.

In my opinion the error rate you previously reported is way excessive, getting 1000s even in a single hour, however if the line is stable when interleaved, from a dumbed down view point openreach may just say its working as described and then refuse to do anything.  So aaisp will need to keep pushing them probably.

Did you tell aaisp about what you was told by your installation engineer?

I know, problem is I noticed I got blips of packet loss of 1% occurring every so often on the AAISP CQM when using the HG612. Now I have 'R: 16' (FEC) on the downstream, assuming the HG612 doesn't change that, I could put that HG612 back on perhaps late tonight and try it again. I wish there was a way for Lantiq chipsets to work on MDWS as they just seem to work a bit better with my line. On the ECI /r with LEDE on the latest firmware I'm sure I recall getting around 75Mbps down, where on the HG612 I believe it was previously 69Mbps when it was 'R: 0', and of course almost none of the intermittent blips of 1% packet loss showing on AAISP's CQM when using a Lantiq chipset.

Yes I told AAISP about it, and I also told the engineer today but the engineer (who I found out was an SFI engineer) said if there was a problem with water as the previous engineer mentioned then there would be a noticeable issue on the voice side. Given the periodic bad weather here it should've been noticeable. However he did say that it might be there's no or very few spare pairs which are working correctly and could be why they mentioned it as it is something that still needs doing but isn't affecting your pair at the moment.

Here's my latest stats ( https://pastebin.com/raw/SrJV1xU9 ), and I only think it's thanks to having 'R: 16' for some reason on the downstream that I'm not seeing lots of error seconds again and mostly FEC errors instead even though I'm currently on fastpath.

Here's the bitloading chart from the DrayTek too: https://i.imgur.com/03i4kck.png

Ignore the bin numbers below the charts, they don't quite line up for some reason.

AAISP appear to want to keep the fault ticket open for now and let more time pass to get more statistics. Last I checked on the control panel they're scheduled to discuss this with me again on or by Monday.

EDIT: Oh, forgot to mention that the engineer didn't seem too keen for me to use the ECI /r or HG612. He called them old and probably don't sync as well as more modern modems do. I did mention however that they were used as more proof that it's not my own equipment.
« Last Edit: October 20, 2017, 06:00:33 PM by Ixel »
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #68 on: October 21, 2017, 01:27:54 AM »

A brief update tonight to say it appears the problem is still there.

I switched to the HG612 and immediately I got lots of error seconds accumulating so quickly unplugged it. Interestingly it was 'R: 0' too, so no FEC errors. The 'R: 16' seems to be unique to the DrayTek, or possibly a modem with a Lantiq chipset. The unlocked ECI /r with LEDE also started getting many errors but I can't see what the R / RFEC was sadly. It was clearly more than 0 because it was also accumulating FEC errors in addition to CRC errors. I've got brief stats from both before disconnecting them and will report my findings to AAISP in the morning, for now I've left a pin on my CQM to explain what the downtime was about in case someone in AAISP wonders before I get to explain what happened. I'll also post the stats here tomorrow most likely via an edit to this post, so keep an eye out, it will be in the morning.

DrayTek Vigor 2860Vac is currently the only modem/router that's keeping errors in check, though the CQM and ping graphs are showing some odd spikes every so often which I'll also post here tomorrow morning.

I think the next logical step might be to either temporarily buy or hopefully maybe borrow a recommended modem/modem+router combined from AAISP for me to run on the line to see what the results are from that. I highly doubt both the ECI /r and the HG612 have a fault though, just doesn't seem very likely for two different devices to perform similar in regards to CRC errors. I'll see what they say however, but with this discovery I have a feeling the fault will remain open even if the DrayTek is the only modem so far that appears more stable on the DSL signal.

In the meantime, good night, and it looks like the problem isn't over. I'm guessing the tester's modem the engineer was using yesterday morning was as efficient as the DrayTek Vigor 2860Vac is for handling the errors (which is a bit unfortunate too).
« Last Edit: October 21, 2017, 01:39:27 AM by Ixel »
Logged

Chrysalis

  • Content Team
  • Addicted Kitizen
  • *
  • Posts: 7382
  • VM Gig1 - AAISP L2TP
Re: My line might have issues?
« Reply #69 on: October 21, 2017, 03:04:21 AM »

you probably best keeping it plugged in, if its bad enough to show packet loss, then aaisp can use that as evidence to give to openreach.  In other words its better to have worse conditions when reporting a fault.
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #70 on: October 21, 2017, 09:23:46 AM »

you probably best keeping it plugged in, if its bad enough to show packet loss, then aaisp can use that as evidence to give to openreach.  In other words its better to have worse conditions when reporting a fault.

Good point, I'm going to ask this morning when mentioning what I've discovered last night and see what they suggest. As I say I feel it's highly unlikely for both the ECI /r with LEDE and the HG612 to both be faulty. The Openreach SFI engineer wasn't too keen on the fact I was using a HG612 or ECI /r, classifying them as old technology, so I'm wondering if the HG612 was used as proof whether it would just allow the engineer to blame the HG612 as being the faulty device. It's possible AAISP might suggest to buy or perhaps loan a specific test modem to eliminate all doubt. As long as it's a Broadcom chipset I feel it should be certain to prove there's problems on the line. The other problem is the engineer's tester seems to be acting similar to the DrayTek, in that it's handling the errors (mostly making them into FEC errors although fastpath).

I'm just about to email them now, I'll post the brief stats here shortly as well.

EDIT: Statistics included.

HG612 stats: https://pastebin.com/raw/ug98KYqT

ECI /r with LEDE firmware: https://pastebin.com/raw/6sqkU8am

Here's the DrayTek stats since last night: https://pastebin.com/raw/TYR3kE0R

--

EDIT (2017-10-21 13:27): Had an update from AAISP that TT said to them that they don't think there's anything wrong. Mostly understandable if I'm using the DrayTek. I've been told that this probably can't be progressed further until Monday, but I also haven't been asked yet to use the HG612 and collect data myself, pretty much just to wait by the looks of it. I guess best thing to do is wait then and hope Monday brings something new and progressive. I could use the line with the DrayTek but it just concerns me it's the only modem I can pretty much use now in order to have a reasonable connection.
« Last Edit: October 21, 2017, 01:34:09 PM by Ixel »
Logged

ejs

  • Kitizen
  • ****
  • Posts: 2078
Re: My line might have issues?
« Reply #71 on: October 21, 2017, 05:01:58 PM »

Regarding the apparent difference in performance between the Draytek and the ECI modem running LEDE:
See http://forum.kitz.co.uk/index.php/topic,18511.0.html
(Although one oddity is that the value 8 in the script from the Netgear DM200 for the operator select command doesn't seem to match the value for BT in the source code.)

Perhaps the Draytek firmware specially configures the Lantiq DSL firmware for Openreach VDSL2, whereas LEDE does not.
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: My line might have issues?
« Reply #72 on: October 21, 2017, 05:59:56 PM »

Regarding the apparent difference in performance between the Draytek and the ECI modem running LEDE:
See http://forum.kitz.co.uk/index.php/topic,18511.0.html
(Although one oddity is that the value 8 in the script from the Netgear DM200 for the operator select command doesn't seem to match the value for BT in the source code.)

Perhaps the Draytek firmware specially configures the Lantiq DSL firmware for Openreach VDSL2, whereas LEDE does not.

Interesting, you could be right about that. Same probably applies for the engineer's tester's modem.

I've lost sync twice today due to heavy rain, if it rains heavy again and drops then I'll quickly run a PSTN test but I bet it passes. I tried it not long ago and it passed. It can't be normal for VDSL2 to completely drop with heavy rain surely?
Logged

Chrysalis

  • Content Team
  • Addicted Kitizen
  • *
  • Posts: 7382
  • VM Gig1 - AAISP L2TP
Re: My line might have issues?
« Reply #73 on: October 21, 2017, 06:40:56 PM »

dropouts shouldn't be happening no
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: My line might have issues?
« Reply #74 on: October 21, 2017, 06:44:17 PM »

I've lost sync twice today due to heavy rain, if it rains heavy again and drops then I'll quickly run a PSTN test but I bet it passes. I tried it not long ago and it passed. It can't be normal for VDSL2 to completely drop with heavy rain surely?

"No" is the one word answer to your question. I believe that you mentioned, earlier in this thread, something about a wet joint? I suspect that the circuit is borderline in terms of usability.

The current A&A modem of choice is now the ZyXEL VMG1312-B10D, which they are selling for £60.  :-X
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.
Pages: 1 ... 3 4 [5] 6 7 ... 11