High packet loss and other DNS issues

[#1]

After being better for a few weeks mine is back to packet loss every evening.

Seeing no problems on the route to ThinkBroadband:

 Host                                        Loss%   Snt   Last   Avg  Best  Wrst StDev
 1. gw                                        0.0%    80   12.8   4.9   1.6  12.8   1.5
 2. ???
 3. 141.hiper04.sheff.dial.plus.net.uk       87.3%    80   11.0  19.1  11.0  53.2  15.8
 4. be3-3106.psb-ir02.plus.net                0.0%    80   12.4  17.7   8.1  64.1  12.8
 5. 195.99.125.142                            0.0%    80   10.2  19.0  10.1  57.4  11.4
 6. 213.121.193.207                           0.0%    80   12.0  13.8   8.3  32.1   4.8
 7. ???
 8. te2-1-9.star10g.bdr-rt3.thdo.ncuk.net     0.0%    80   13.7  16.8   9.1 116.2  13.7
 9. po4-31.core-rs2.thdo.ncuk.net             0.0%    79   18.1  21.6  10.8 105.0  17.4
10. pingbox1.thinkbroadband.com               0.0%    79   16.6  17.4   8.2  66.6  12.8

[#2]

Just dropped in for an update, and yet again, surprise surprise, still issues.

Why should customers be spending all this time do Plusnets job for them?

Surely it should be obvious on their systems there are issues?

New network to help with capacity?

Yeah right. Cant even get that right.

What an embarrassment.

[#3]

I think some people just don't appreciate how difficult some problems can be to track down.

[#4]

As a telecoms company though, you would think there would be some kind of logging on the network to find problems. Seems a little old school to rely on thinkbroadband forum charts.

[#5]

Started earlier to

Also got a US server to ping me for 24 hours not showing near as much loss 

[#6]

Personally, I haven't been having issues the past few nights. Teamspeak has been fine, Twitch has had no buffering. I'll keep an eye on this. Can't do a monitor as the monitor just bugs and shows 100% packetloss all the time.

[#7]

My packet loss has returned today after resyncing my modem and changing my IP (still in the 52 range).

Month of historical data here

 You'll note there is a massive chunk of packet loss on the 4th of August between 3am and 4am on the day this started. What was changed on the 4th?

[#8]

I dropped my PPP session earlier for the first time in ages and I'm now seeing this which looks strikingly familiar :

What's surprising here is that I'm routing through Colindale (PCN), whereas the majority of those reporting problems up until now have been routed through Southbank (PSB).

Tracing route to pingbox1.thinkbroadband.com [80.249.99.164]
over a maximum of 30 hops:

  1     3 ms     3 ms     3 ms  home.gateway [192.168.1.254]
  2     *        *        *     Request timed out.
  3     *        *        *     Request timed out.
  4    15 ms    14 ms    15 ms  be3-3102.pcn-ir02.plus.net [195.166.143.132]
  5    14 ms    13 ms    13 ms  195.99.125.140
  6    14 ms    14 ms    13 ms  195.99.127.83
  7     *        *        *     Request timed out.
  8    16 ms    16 ms    15 ms  te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
  9    17 ms    14 ms    16 ms  po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
 10    14 ms    13 ms    13 ms  pingbox1.thinkbroadband.com [80.249.99.164]

Trace complete.

Save

Save

[#9]

@bobpullen Well, I'd implore you not to restart your PPP session again until this is resolved for all customers. Plus net now have access to my connection from the customer side (@dave) which is currently showing the low-level packet loss, and I'm sure you can provide testing/equipment to debug the high-level packet loss.

[#10]

"I think some people just don't appreciate how difficult some problems can be to track down."

Well I think we can all appreciate that the fact PN have not really kept people informed on what this issue is until a horde of users flowed in demanding explanations, and their support is all over the place with phone line support and web support responding with "It's an issue on your end", probably written on the back of their hand. Also, many people have been paying for sub-par broadband for 6 weeks now. I think those people should be shown a bit of sympathy back in some respect from PN other than just "sorry, being looked into, deal with it" week on end.

Oh, and my monitor, set a stopwatch, wait until past 10am, watch as internet becomes degradingly worse for each day.

[#11]

@bobpullen Welcome to the packet loss club 

I do believe that I had issues with DNS back in July through Colindale.

Today's traceroute:

Tracing route to ntp.plus.net [212.159.13.50]
over a maximum of 30 hops:

  1     1 ms     1 ms     1 ms  router.asus.com [192.168.1.1]
  2     *        *        *     Request timed out.
  3     *        *        *     Request timed out.
  4    13 ms    13 ms    14 ms  be3-3102.pcn-ir02.plus.net [195.166.143.132]
  5    14 ms    15 ms    18 ms  be1.pcn-ir01.plus.net [195.166.129.160]
  6    13 ms    13 ms    12 ms  cdns02.plus.net [212.159.13.50]

Tracing route to pingbox1.thinkbroadband.com [80.249.99.164]
over a maximum of 30 hops:

  1     1 ms     1 ms     1 ms  router.asus.com [192.168.1.1]
  2     *        *        *     Request timed out.
  3     *        *        *     Request timed out.
  4    28 ms    13 ms    15 ms  be3-3102.pcn-ir02.plus.net [195.166.143.132]
  5    14 ms    13 ms    13 ms  195.99.125.144
  6    13 ms    14 ms    13 ms  195.99.127.87
  7     *        *        *     Request timed out.
  8    23 ms    13 ms    13 ms  te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
  9   286 ms    54 ms    16 ms  po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
 10    14 ms    15 ms    14 ms  pingbox1.thinkbroadband.com [80.249.99.164]

Trace complete.

http://usertools.plus.net/@gateway/

Sorry, the tool failed to identify your gateway; this does not mean that anything is wrong with your connection.

Posting the following information as a reply to the Gateway Checker Issues topic in the Community forums will help us to diagnose the problem.

• core2-hu0-18-0-0.colindale.ukcore.bt.net (62.172.103.16) 5.390 ms
• *
• *

[#12]

---

Riza wrote:

http://usertools.plus.net/@gateway/

Sorry, the tool failed to identify your gateway; this does not mean that anything is wrong with your connection.

Posting the following information as a reply to the Gateway Checker Issues topic in the Community forums will help us to diagnose the problem.

• core2-hu0-18-0-0.colindale.ukcore.bt.net (62.172.103.16) 5.390 ms
• *
• *

---

It's been known for a while that some of the inbound interfaces on the IR's are non-responsive to ICMP, which is what's causing the identification failure here. Whilst that in itself shouldn't be causing problems, it would be interesting to see if others that are seeing this packet loss are also on one of these routes? Perhaps a few people could try visiting the gateway checker to prove/disprove this theory.

I'm still on the 'bad' PPP session and my gateway can't be identified either.

Packet loss starting early today too:

[#13]

Telling me what Gateway

Array
(
    [1] =>  7  core1-hu0-9-0-1.colindale.ukcore.bt.net (195.99.127.18)  4.681 ms
    [2] =>  8  195.99.125.141 (195.99.125.141)  5.211 ms
    [3] =>  9  *
    [4] => 10  243.27.6.51.dyn.plus.net (51.6.27.243)  22.769 ms
)

You are currently connected to gateway pcn-ir01.

This is located in Colindale Point of Presence.

This is my TBB today

This was it on the 23 August

Oh I long for the good old days

[#14]

Thanks, that's that theory out of the window then.

I do notice that the loss seems selective, certainly when checking the outbound route e.g. I get a lot of loss to www.google.com but relatively little to google.com which resides at a different IP.

Some routes I get zero loss to e.g. here (community.plus.net).

[#15]

@bobpullen

FWIW.  Way back in June/July when you started to move people to the "new" network.  The gateway checker showed me as something ir01 as I recall.  Speed was fine, 70 down, 19 up - even though I was still on 40/20 at that time.  No packet loss.

Then when you moved me to 55/10 and applied the cap, the problems started and gateway checker could no longer identify the gateway.

Now I'm on a static IP (with packet loss largely resolved) and the gateway checker shows btw-bng01.

If it means anything at all, I did notice that after the 55/10 cap was applied most speed checkers showed 55/10.  Except dslreports.  That repeatedly showed 55/19.  I checked my local monitor and it really was moving data to dslreports at 19.

Some routes on your network not applying the cap properly?  On the static IP, everything now shows up as 9.5.