198.18.1.x address problems

[#1]

The most commonly automatically assigned DNS servers, 212.159.6.9 and 212.159.6.10, and also 212.159.13.49 and 212.159.13.50, are all Plusnet's. The DNS servers are owned by BT in the sense that Plusnet is owned by BT.

There are two different ways the change the Technicolor router's DNS settings, one way bypasses it's internal DNS server and so would actually avoid the problem even if you configured it so that its DHCP server gives out 212.159.6.9 and 212.159.6.10 as the DNS servers. I think the other way can add additional DNS servers, but the automatically assigned Plusnet ones will still be listed and may get used sometimes.

The part you quoted doesn't really support what you said, since changing the computers to use the DNS servers directly avoids the problem, as does using a non-Technicolor router, which wouldn't be the case if the DNS servers themselves were the problem.

[#2]

Hi,

Sorry to resurrect this topic, I am looking at a computer for my father in law that has been experiencing these same issues.

I took it off DHCP and added the preferred and alternate DNS and that seems to have sorted the issue for now. Looking at the extensive posts is the "official" fix still to telnet into the router and turn spoofing off?

Seems a bit extreme and not  an easy thing to do for your average joe I cant see any word from plusnet as to providing new router or step by step guide to how to telnet and fix.

Thanks

[#3]

---

@bibbypeterpn wrote:

Sorry to resurrect this topic, I am looking at a computer for my father in law that has been experiencing these same issues.

It would probably be a good idea to exactly describe the symptoms you are now seeing after changing the DNS settings on the PC. Once you have changed the DNS settings, this problem cannot affect you - so either the DNS settings haven't been picked up or you have a similar but slightly different root cause.

[#4]

I am experiencing no issues after following the advice and amending the DNS details on the local PC.

BUT I was enquiring what plusnets "official" suggestions are for the route issue.

I am concerned my current "fix" only fixes the current PC. If we added a computer onto the network we would have to do the same fix or it would also exhibit issues.. so I was going to look into the router "fix" but considering the age of this topic wanted to check this was still the recommended approach.

[#5]

Ah, sorry, I misread.

As far as I know the official recommendation is still the same. I sent Bob some traffic dumps last year and he was looking into something further after a Windows 10 update started to cause issues, but I haven't heard of any conclusion.

Personally, I'd dump the router as overall it seems bad for the router to be doing anything even remotely like this - perhaps if you call plusnet up and explain the issue they might offer to send you a more recent router? (I think they ship plusnet branded BT hubs to new customers these days?)

[#6]

Thanks for the information

Ill tell him to get onto them as his contract is up this month

[#7]

Rashly or otherwise I'd like to comment further on this, in the light of experience. And it might still be puzzling some new Plusnet subscribers.

I too suffered from this 198.18.1.x address problem with my Technicolor TG582n router, put up with it for a long time and rebooted the router when needed. I got fed up of this and wanted a solution, so I tried several suggestions from these boards, put up with it a bit more and then applied the "WANDownSpoofing=disabled" change to the router.

This did indeed solve the return of 198.1.18.xxx addresses by the router, but not the underlying issue which was presumably causing the WanDownSpoofing page to be returned by the 582n. Instead I just got periodic loss of contact with the Internet which eventually would right itself. This is how it remains.

Previously I thought I had noticed an association with rising FEC errors (as reported by DSLstats) with this issue. It occurred again today and now I do feel they are related (or simple line errors causing loss of contact with the Internet?). High FEC errors, loss of contact with the Internet, hence loss of contact with the DNS and the rest follows.

Now the usual reply is: "FEC errors don't matter, they are just Forward Error CORRECTION"

My reply to this is: "Yes. But you only get Error Correction when there are errors to correct!"

ADSL is not perfect, so there will always be FEC errors even when everything is working well. But, they can sometimes rise alarmingly - remember, the DSLstats chart for FEC errors is logarithmic, so a bit of an upturn on the plot can mean orders of magnitude increase. My recent drop off was accompanied by a downstream FEC error change from a normal rate of 3000 to 100,000 FEC per minute.

As to what is causing the increases in FEC error rates, that is another matter, though in this case they were accompanied by a similar rise in downstream CRC errors, from around 3 to a peak of 1000 CRC per minute. The average errors for CRC/FEC errors were 423 and 285,812 per hour downstream. No HEC errors. These rise in FEC error rates do seem to occur now and then, for no obvious reason (not always accompanied by CRC errors).

[#8]

@pvmb wrote:

Rashly or otherwise I'd like to comment further on this, in the light of experience. And it might still be puzzling some new Plusnet subscribers.

---

It puzzles some old ones too.  It appears that a new release of Outlook (Office 365) is now tripping over this fault in the TG582n router. Office on my machine went up to level 16.0.8625.2127 on 17/11/17, and later that day I found that all my emails were downloading (to Outlook) extremely slowly (average one email a minute approx.).  The problem did not arise when I used tethering to my mobile phone in place of the PlusNet broadband. The problem persisted until with help I was directed to this thread, albeit in connection with a different problem.  I added a substitute DNS server in my TCP/IP configuration (as a temporary workaround) and the speed has gone back to normal, perhaps even faster than it was before. I have no idea what has changed in Outlook that now trips over this fault in the router software, but trip over it, it does.  I have not as yet tried reconfiguring the router as detailed in msg 103, but may do so later.  PlusNet should have made sure everyone got a router fix for this. @bobpullen indicated earlier that this was not likely.

My wife's computer on the same router strangely does not seem to have been affected, but I have a further trial to do there since it has only recently updated to the latest Office version.

@Kelly, Ref the thread where I reported this (https://community.plus.net/t5/Broadband/POP3-emails-throttled-on-PlusNet-broadband-affects-both-PN-a..., please note the above info re Office, as this is likely to come up for other users with this router and Office 365, and it is not at all obvious as the source of the problem.  I spent ages trying to find the reason on Windows and Office.  What does surprise me is that Tech Support did not come up with this, and that they merely tweaked my line parameters. I will go back to them, waving (but not waiving) my ticket.

[#9]

AtP,

Don't be too harsh on the front line support staff. 

This problem is not an obvious culprit ... unless you've seen it before and consider all possible causes of peculiar computer behaviour ... remember I started this topic and IIRC it took a year to get a profile on the issue.  This is not a finger print I would expect the first line tech guys to consider.

[#10]

Yeah, of course you're right. I'm not a tough nut, but I do grumble a lot.

But, given the new release of Office and the number of these routers still out there (whatever that is), there may be a lot of people now coming across this problem, so Support had better be made aware of it. or customers will get a useless 'fix' like I did, and you have no idea how much time it has cost me to sort out.  Well, yes, you probably do, because I see elsewhere that your inclination, like mine, is to stick at something to find the root cause, even when a fix is available that avoids the problem.

That reminds me: I notice on related threads that many people refer to a "route cause".  I doubt a pun was intended, but it was apposite here!

[#11]

@Attila_The_Pun, just replied to the thread over here.

[#12]

I guess this topic is considered well past its sell by date, however for anyone interested, may I draw your attention to the following thread: https://community.plus.net/t5/Broadband/PPP-Drops/td-p/1537570

In particular from post 5.

After manually applying the fix described in this thread to my TG582n router I had no repeat of the WANDown address spoofing problem until recently. I then found, despite WANDown spoofing still being disabled, the router somehow had acs.plus.net (the Plusnet auto configuration server) listed in its address spoofing table and I had to rerun run the router command dns server debug spoof clear to remove the address from the router spoofing list (then saveall exit). Everything then returned to normal.

I don't know how, despite WANDown address spoofing being disabled, the router still listed this address. Presumably it could happen again?

[#13]

Yet again this somewhat points to an underlying intermittent DNS failure.