Print

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

[BobC]

It kinda depends.  If rebooting the router fixes it but you sync at more-or-less the same rate and configuration, I'd be thinking there is a problem with the line card.

However if rebooting fixes it and your sync rate is a lot lower, then its likely interference from something.

If I catch it in the moment when both the corrected DTU's and uncorrected DTU's are going wild causing these ES, SES and CRC errors to jump to crazy numbers and I restart the router it resolves it. The INP sometimes changes from 51 to 52 or vice versa and my gateway used to be cor1 Lon or something but in the past month the gateway as changed to Man-RE0 or the London equivalents. Sync rate remains the same and those correction numbers fall dramatically until the next encounter. The Fritz!Box 7530 router I had before the replacement arrived had the same issue but there seemed to be a very clear pattern to it with it happening once per month of uptime. With the replacement router it can be 0 errors for a day, a week, 2 weeks then a flood of errors come in due to the dramatic rise in corrections. The per minute corrected averages used to settle around about 18 on the old router until that monthly spike. Now though the Corrected DTU's are just bouncing allover the place with no clear pattern. I just took a look at the router stats right now and those corrections are rising once again. The last time I left it alone without rebooting the dlm intervened due to the error count exceeding the daily limit but 2 days later I was back from 4dB to 3dB. Should I just leave it alone and let these error counts rise? It hasn't completely fell of a cliff as of yet but with 32k corrections and some errors forming in the past 15 minutes alone it's only a matter of time. I guess the issue is outside and beyond my control because I changed the dsl cable, the router and the power adapter this time around and it still happens, admittedly much more often, but nothing looks out of the ordinary or changes by much when checking the limited stats provided by the router such as signal to noise ratio, the spectrum information, the Fritz!Box cpu usage and temps etc.

As for the router dropping the PPP session every 2 minutes when saturating the line since fitting this new cat7 cable, any idea what that issue could be? Everything seems to point to a faulty cat7 cable or some kind of interference knocking the connection out temporarily between the router and ISP. We will try removing that cable first of all and hopefully that resolves it. If not I have no idea how we'll go about fixing this new issue.

[burakkucat]

As for the router dropping the PPP session every 2 minutes when saturating the line since fitting this new cat7 cable, any idea what that issue could be? Everything seems to point to a faulty cat7 cable or some kind of interference knocking the connection out temporarily between the router and ISP. We will try removing that cable first of all and hopefully that resolves it. If not I have no idea how we'll go about fixing this new issue.

I would be very suspicious of any cable that is labelled CAT7, especially one with a flat profile. Proper CAT7 (and above) cables are really for data centre usage. In a typical domestic environment CAT5e or, if you must have a bigger number, CAT6 be perfectly adequate. I'll hazard a guess that our member "Reformed" will probably be able to state actual bandwidth v distance figures for the various category cables.

[Alex Atkin UK]

I would be very suspicious of any cable that is labelled CAT7, especially one with a flat profile. Proper CAT7 (and above) cables are really for data centre usage. In a typical domestic environment CAT5e or, if you must have a bigger number, CAT6 be perfectly adequate. I'll hazard a guess that our member "Reformed" will probably be able to state actual bandwidth v distance figures for the various category cables.

Yeah I only made that mistake once, it wouldn't even do 10Gbit.

I do have a CAT8 flat though, wanted to over-spec as its running from the PoE switch to an Ubiquiti PoE converter for my Litebeam WiFi link across the street.  So wanted to be absolutely sure there was plenty of grounding on that cable in the hope a nearby lightening strike wouldn't take out the switch.

[BobC]

So I removed the cat7 cable from the router port tried to download something from Steam on the downstairs PC with a previously stable ethernet cable and I lost connection 3 times in a row with Zen. Event log attachment available below. I downloaded CyberPunk 2077 on it last month which was 64gb and all was good. I downloaded Mass Effect Legendary Edition on the upstairs PC which was like 105-110gb with that temporary cheap cat6 cable that was being ran up the stairs and that was fine.

So today I was downloading Prey on the downstairs PC which was like a 20gb file. 3 drops occured early doors then it was fine. Re-inserted the cat7 cable and things continued to look stable for the rest of the download. Went back upstairs and CyberPunk had a 4 or 5gb update so I started downloading that through the cat7 cable and the connection with the ISP provider remained stable, there was a slight wobble though. The connection speed dropped to 6kbs for a few seconds then shot back up to 66Mbps. I checked the dsl information tab and spotted this Min. effective data rate kbit/s 6. Was previously at 60Mbps. Never seen this issue before either... I think I need to get shut of FTTC the issues I have with it stresses me out. Unfortunately I didn't have a ThinkBroadbandMonitor running in the background to test the amount of packetloss...

The issue with the high rise in Corrected DTU's which in turn causes errors on the line until a router reboot definitely seems like something outside. I can't keep on blaming the FritzBox routers and the cables provided surely? It's a complete random occurence. Corrected DTU per minute average 10. Last 15 minutes 23 with a 2 day 7 hour uptime. I was going to switch back the FritzBox provided dsl cable back to the one provided with my BT HUB years ago but it's a lot of messing about.

The few errors that accumulated on the dsl information screenshot occured when messing around with the ports on the router and undoing the cable tie. I don't understand the spectrum stats but I included the minimum / maximum graph even though I have never seen them stoop that low whenever I have checked that graph out with the min / max unticked. The signal to noise ratio graph also always stays between 3 and 4.

[burakkucat]

Your problem(s) is (are) very perplexing.    I do not understand why the circuit behaves as you report.

[Alex Atkin UK]

Honestly I'd be thinking the router is faulty somehow.

[Edinburgh_lad]

I believe your problem is similar to mine. I don't think this is to do with your Fritzbox (I get the same results on my Zyxel and Draytek, in addition to my Fritzbox).

In my case, there is a high resistant fault (something to do with jelly crimping) that openreach won't fix because of where the cables are and how accessible they are. I believe there's also noise on the line (it seems that when our neighbours are in, and they are not in often nowadays, and when they put something on, presumably the boiler or water tank heater, for every 2 errors on the line (ES on fritzbox), I usually get one SES, but this, I believe depends on how intensive the noise thing is, which makes me think it's a hot water tank). During this the min effective data rate also gets reduced to 3kbps).

So far, I've been trying to resolve our issues for over a year now since we started working from home and switched over to VDSL. Lots of testing, lots of money spent, too, lots of emails to Plusnet and lots of openreach visits (14 at least). Openreach weren't transparent with us: two mentioned a high resistance fault, whereas another admitted they didn't know where the interference was coming from, but because no-one complains from the neighbours who also have the problem as confirmed by them and BT wholesale (all retired and don't use the internet as much), they'll not investigate just one flat.

[Weaver]

@Edinburgh_lad One extreme technique that I employed last year was to order an additional line and then cancel (‘cease’ ?) the bad one.

[tubaman]

I don't believe this is a physical line issue as only the PPP connection appears to be dropping. More likely either a Zen config issue or an issue with a linecard.

[Black Sheep]

I believe your problem is similar to mine. I don't think this is to do with your Fritzbox (I get the same results on my Zyxel and Draytek, in addition to my Fritzbox).

In my case, there is a high resistant fault (something to do with jelly crimping) that openreach won't fix because of where the cables are and how accessible they are. I believe there's also noise on the line (it seems that when our neighbours are in, and they are not in often nowadays, and when they put something on, presumably the boiler or water tank heater, for every 2 errors on the line (ES on fritzbox), I usually get one SES, but this, I believe depends on how intensive the noise thing is, which makes me think it's a hot water tank). During this the min effective data rate also gets reduced to 3kbps).

So far, I've been trying to resolve our issues for over a year now since we started working from home and switched over to VDSL. Lots of testing, lots of money spent, too, lots of emails to Plusnet and lots of openreach visits (14 at least). Openreach weren't transparent with us: two mentioned a high resistance fault, whereas another admitted they didn't know where the interference was coming from, but because no-one complains from the neighbours who also have the problem as confirmed by them and BT wholesale (all retired and don't use the internet as much), they'll not investigate just one flat.

Methinks you are jumping the gun, slightly.

If your circuit has a high-resistance fault (HR), then that absolutely requires mending first before you can even contemplate the second scenario you are alluding to, which is REIN.

The metallic path facility (MPF), has to pass a various suite of tests to ensure its integrity is capable of eliminating normal levels of noise interference'. If there is a HR prevalent, then the circuit is highly susceptible to these normal levels of noise, causing multiple errors/LOS/PPP drops.

Once the MPF is deemed as good, only then can the finger of suspicion be pointed towards REIN/PEIN/SHINE events should the issue continue.

Plus, OR will of course investigate - whether it's one flat or 200 flats. The technology has changed indescribably in the last 3-4yrs for pin-pointing patterns of noise, timings, circuits affected etc etc ..... but REIN faulting is still a gratis service provided by OR, it is not a given. If Mrs Jones powerline adapter in the flat above yours is causing high frequency noise events, OR nor your ISP can be held accountable for that.

Regardless of that though - if you suspect you still have a HR, why not try and force the fault by using your landline on 'quiet line test' or ringing someone and see if your error rates continue to climb. That is a sure-fire way of knowing.

[Moderator edited to remove one spurious [/b] tag.]

[Reformed]

I'll hazard a guess that our member "Reformed" will probably be able to state actual bandwidth v distance figures for the various category cables.

Not really! Those are minimum standards they've passed but most of the time the shorter runs we have in our homes will handle way more.

I use 7 for short runs as it's a more durable cable. It is prone to more faults when being made though. It's not fun to attach the RJ45  plugs to.

That's what I do anyway. 7 for very short patch cables, DACs for sub-5m unless I have to use RJ45 in which case 6 or higher for 10G just to be sure, 5e for 2.5/5G, fibre for anything longer. I buy premade so that they may be returned and don't make enough cables to justify a drum as I don't use that much of those types. Freebies with equipment cover most of my needs.

Pluggables, so SFPs, don't seem to transmit at as high a power as direct ports in some cases so don't run them at 10G over more than 30 metres.

With 10 metre runs most Cat5e and above should be fine with 10G, with 6 certified to 55 metres and 6a and above 100.

[Edinburgh_lad]

Methinks you are jumping the gun, slightly.

If your circuit has a high-resistance fault (HR), then that absolutely requires mending first before you can even contemplate the second scenario you are alluding to, which is REIN.

The metallic path facility (MPF), has to pass a various suite of tests to ensure its integrity is capable of eliminating normal levels of noise interference'. If there is a HR prevalent, then the circuit is highly susceptible to these normal levels of noise, causing multiple errors/LOS/PPP drops.

Once the MPF is deemed as good, only then can the finger of suspicion be pointed towards REIN/PEIN/SHINE events should the issue continue.

Plus, OR will of course investigate - whether it's one flat or 200 flats. The technology has changed indescribably in the last 3-4yrs for pin-pointing patterns of noise, timings, circuits affected etc etc ..... but REIN faulting is still a gratis service provided by OR, it is not a given. If Mrs Jones powerline adapter in the flat above yours is causing high frequency noise events, OR nor your ISP can be held accountable for that.

Regardless of that though - if you suspect you still have a HR, why not try and force the fault by using your landline on 'quiet line test' or ringing someone and see if your error rates continue to climb. That is a sure-fire way of knowing.

[/b]

I don't want to hijack the OP's post, but in the words of one Openreach engineer that came to see us (in fact the very first one): 'it is what it is', which sums up everything. Seemingly there were no spare 'good' pairs in the cable and the cable is underground and they couldn't access it, so that's where we are.

However, today, Virgin Media have brought their equipment in to put their FTTP in the area, so I can't wait to wave goodbye to Openreach. 

[Black Sheep]

I don't want to hijack the OP's post, but in the words of one Openreach engineer that came to see us (in fact the very first one): 'it is what it is', which sums up everything. Seemingly there were no spare 'good' pairs in the cable and the cable is underground and they couldn't access it, so that's where we are.

However, today, Virgin Media have brought their equipment in to put their FTTP in the area, so I can't wait to wave goodbye to Openreach.

Yeah - I read a lot of that 'He said, she said' stuff.

If you've had a poor engineer, well the odds are that's going to happen with circa 30,000 of them employed by OR .... but you've had "14 engineering visits at least". That sounds more to me like somebody isn't in possession of the full facts, or has misinterpreted what's been said by the engineer, or your circuit is performing within the 'cone of acceptance' agreed between OR and the ISP's ....... either that or there's a conspiracy theory going on between 14 different OR engineers, and your ISP ?.

If for whatever very bizarre reason OR can't access our their own underground cables, there's a plethora of alternative solutions that can be used. Simply put, OR can not leave a live fault on a customers MPF - with the previously mentioned advances in remote testing and collating of data, there's not a cat in hell's chance OR's 14 different engineers have simply walked away from a full on fault condition without either fixing it, or applying remedial measures to fix it.

[Weaver]

And me also not wishing to cut in on the OP’s thread, but what BS just said is relevant to me and my always-fixed faults, but never fixed long-term. See https://forum.kitz.co.uk/index.php/topic,26851.msg450753.html#msg450753 - If you’re willing, perhaps contribute on this topic but in that thread. That would be appreciated.

[BobC]

I'm going to leave this post short with screenshot attachments. A continuous flow of corrections / errors building yet again. From 0 to this after several hours with 5 days 12 hours of uptime this time around. The entire thing is just unstable and unreliable :/. I'm not going to bother rebooting the router to clear them, evidently had enough. FTTC is just rubbish. The issues that crop up on my line or at the cab or wherever the latest issues are never seems to be easily identifiable, sucks it really does.

That peculiar isssue with the downloads is still happening as well, The Fritz!Box 7530 router disconnecting the link with the ISP then re-authorising it, then all of a sudden it's stable? I have noticed that saturating the line with Usenet downloads appears to be far more stable than Steam in regards to PPP disconnects, but as this issue is recent, and we already tried removing the only new thing ie one new ethernet cable, that issue is just as confusing as any other irregularity I've experienced.