Virgin routing issues

[studders]

Thanks for that Mark, that does clear up a few questions I had.
I only have a very basic knowledge of how networks operate (It's all done by magickry right?) and have absolutely no idea how to use the utilities provided in OS X. A bit of bedtime reading may be in order.
If it makes my head hurt I will just put up with the delays. :?

 

[studders]

OK thanks chaps. Have found it in Applications-Utilities-Network Utility. will give it a bash and see if I can spot what you've said.
Thing is though, even if I do spot something, is there anything I can do about it?
Actually.. forget that, I'll just get myself tangled up in knots.
Will experiment with Traceroute first and get back to ya if I spot something I can't suss out.

 

[jlawrence]

Nope, even if you find something you can almost 100% of the time do absolutely squat about it - unless you want to start bouncing through alternative proxies etc.
But, it's not about being able to do something about it. It's all about knowing that it's not your fault - and it can be fun to ring up the support guys just to wind them up a tad.

 

[studders]

it can be fun to ring up the support guys just to wind them up a tad.

I thought that was obligatory, mind you, in this case it would be rather an expensive call.

 

[Eric The Viking]

Thing #2: cable modem behaving 'normally'

Not sure what 'normally is, but okay.

It's a Motorola dartboard -- old but it's been fine (screwed to a beam above the bench I'm fettling planes on at the moment, up in the attic!). I"m getting a normal light sequence: TX + RX connections OK, data flowing and occasional activity.

Thing #3: some servers (such as my mailserver, which is physically nearby) are reachable, when things like the BBC site aren't.

When you say 'physically nearby' how do you mean?

The work one is down the road, physically in central Bristol. Domestic email is from Yahoo (US). Sometimes, during fault conditions, my local (in the house) mailserver's log shows it could reach Bristol, but not Yahoo. Sometimes both have failed together, in periods of total loss of connectivity, including port 80 stuff too.

Thing #4: it comes and goes, and slows down and speeds up (using web pages as a guide).

Could be a link issue, could also be a contention issue. Have (can) you check the stats on the modem to see if you have a high CRC error rate on your WAN interface?

It's broadband cable rather than ADSL, so I wouldn't expect high CRC errors, as there's far more margin/bandwidth available in the 'local end' (it's fibre from the green box in the pavement, literally outside the house next door). I haven't got a local RT entry for the modem's web server at the moment, so I'd have to plug a laptop directly into it to look at its logs.

Thing #5: when it does work, long-distance (e.g. NZ) is significantly slower than normal, but reachable, implying an odd route.

Hmmm, routing can be an odd beast but your only routing concern will be getting from your house into Virgins network, after that it's their internal routing then out onto their country peer, so that would affect many more people than just you.

I think it's at best an area fault, at worst smoke in somewhere like Telehouse! I didn't explain myself very well earlier (crucial evidence about Virgin admitting the problem was omitted!)

You mention tracert, it would be worth getting a tracert of a known problematic site now, while it's working, then get the same again when it goes castors up. This will tell you where any routing issues might be coming in.

Yep. Here's the obvious one:

Tracing route to ukworkshop.co.uk [92.48.118.217]
over a maximum of 30 hops:

  1     *        *        *     Request timed out.
  2    11 ms    19 ms    11 ms  10.45.136.1
  3     9 ms     8 ms     9 ms  osr01azte-v15.network.virginmedia.net [80.192.0.1]
  4    12 ms    11 ms    11 ms  aztw-core-1a-ge-200-0.network.virginmedia.net [80.1.242.1]

  5    21 ms    19 ms    16 ms  winn-bb-1a-as0-0.network.virginmedia.net [213.105.175.157]

  6    19 ms    17 ms    70 ms  brnt-bb-1b-as5-0.network.virginmedia.net [213.105.172.234]

  7    25 ms    28 ms    33 ms  brnt-bb-1a-ae0-0.network.virginmedia.net [213.105.174.225]

  8    15 ms    15 ms    22 ms  nth-bb-b-as4-0.network.virginmedia.net [212.43.162.218]
  9    18 ms    18 ms    20 ms  tele-ic-1-as0-0.network.virginmedia.net [62.253.184.2]
 10   219 ms   207 ms   192 ms  114-14-250-212.static.virginmedia.com [212.250.14.114]
 11    43 ms    29 ms    27 ms  213.133.130.78
 12    24 ms    23 ms    26 ms  bac.bsq3-hex.as29550.net [217.112.81.94]
 13    22 ms    24 ms    22 ms  mayall.calynet.co.uk [92.48.118.217]

Trace complete.

Ignore the first line - I have a deliberately misconfigured DNS server, which is causing the apparent odd timeout. It's a long story, but it's not an actual problem.

The most likely if the routing tables are regularly reconverging is a flapping route/link somewhere causing a routing relationship to flap, probably in 'the cloud' somewhere.

Makes sense to me. Of course, I forgot to mention the important thing: called them on Friday and today and in both cases (when pressed) they admitted they had a problem. First-off it was all the 'restart your modem' nonsense.

Regarding the speed of rebuilding routing tables, I wasn't explaining quite what I meant: It's behaving as if they're repeatedly restarting some bit of kit/cluster/whatever. Anyway, it's been up for four hours so far - longest since the trouble started.

I'll be saving this week's email logs too...

 

[mailee]

Ah, I thougt I might be able to answer this one assuming it was a novice asking for routing advice ala Hitachi, De Walt, Trend etc. This stuff is all over my head so I'll shut up. :?

 

[llangatwgnedd]

Mine as been off and on all weekend, even had to re-boot the modem at one stage.

Never had this prob before.

 

[Sawdust]

I'm with Virgin braodband and can't wait to get out as they are useless. It's fine very late at night (e.g. 3am) or very early morning but for most of the evening it is unusable with download speeds of about 0.5M bits per sec.

They do nothing to help, just tell you to turn the router off for 15 mins and then back on which makes no difference whatsoever. I'm doing my best to get out of my contract early but they really don't care about anything apart from taking money.

As soon as I possibly can get out I will but if anyone is thinking of joining Virgin just DON'T!!

I understand their cable service is great but their broadband is rubbish.

Cheers
Mike

 

[jlawrence]

It could well point to a problem within telehouse - as my traceroute might as well.
I'd feel fairly happy to say that your traceroute points to a private peering problem. Which would make sense given that you say it keeps bouncing up and down. The private peer drops and the router drops back to routing via a transit and everything comes back up, when the private peer comes back on line it takes priority again. If it happens often then there is something wrong with Virgin's router configs - a flapping peers' priority should be dropped automagically until it's been stable for a defined length of time.

 

[bodgermatic]

Low signal level to blame then I guess and not nothing

Strangely enough, too high a signal level The head end was shouting at your router / settop box, so he added an attenuator to quieten it down. They usually have a dB printed on them somewhere.

 

[Eric The Viking]

It could well point to a problem within telehouse - as my traceroute might as well.
I'd feel fairly happy to say that your traceroute points to a private peering problem. Which would make sense given that you say it keeps bouncing up and down. The private peer drops and the router drops back to routing via a transit and everything comes back up, when the private peer comes back on line it takes priority again. If it happens often then there is something wrong with Virgin's router configs - a flapping peers' priority should be dropped automagically until it's been stable for a defined length of time.

I bow to your superior wozzaname on this (haven't myself looked at hop counts etc. for donks - prob. about the time IPX was the 'protocol-of-of-the-month' centrefold in the Orem/Provo Gentleman's Companion). I didn't know much to start with, and I think I've forgotten that now too, good'n'proper.

Today I have mainly been tradeshow-standin (and me feet urt!), since around 0830, and only just got back to me desk. It looks as though broadband's been reasonably OK all day (haven't ploughed through all the logs - life's too short).

I'm hopeful it's now fixed...

 

[Paul.J]

Yes same here.It seems to be fine today

 

[Sawdust]

Still useless here!

0.6M bits per sec download. It usually speeds up about midnight. They are even telling me I have to pay for the priviledge of leaving!

Cheers
Mike