Kitz Forum
Broadband Related => ISPs => Topic started by: broadstairs on May 23, 2015, 06:28:32 PM
-
I have noticed today that some websites are taking a long time to load, just like it is losing packets, a reload of the page seems to load it OK but it was not DNS as it had already started accessing the page. This seems to me not unlike the issue we had on TT a while back.
Anyone else seeing this?
Stuart
-
Yes, I believe that there is something "not quite right" with TalkTalk.
Earlier today I discovered that my PPPoA session had dropped and restarted but the CPE - CO synchronisation state was undisturbed. Attempts to load certain web pages was just impossible yet others loaded without problem. As I am using a Huawei HG622, I have no means of dropping and re-establishing the PPPoA session so I had to resort to taking down the xDSL layer. It took three iterations of --
xdslcmd connection --down
<wait two minutes>
xdslcmd connection --up
before I regained a usable service. At to the reason, I know not why. :no:
-
On my ADSL connection I did notice earlier that some sites were taking a very long time to load; I'm using the Google DNS for what it's worth.
Doesn't seem too bad now though.
Ian
-
Yes, I also use Google's public DNS servers. I don't believe that the problem was DNS related. :-\
-
I have issues also. On TalkTalk Fibre here.
The downspeed throughput went down from 40Mb to 11Mb for around an hour until I forced a new connection which seemed to fix things. Ping seems a bit higher than normal. Some websites are taking a bit of time to load (thinkbroadband being one).
-
The symptoms of the problems I see are very similar to an issue we had on TT in February 2013 where some packets got dropped usually after the browser had done its DNS thing and contacted the site. Seems to come and go a bit and is worse late afternoon/early evening. I have raised a thread on the TT forums in the hope (more than expectation) that someone will take a look at it tomorrow - we'll see.
Stuart
-
Just taken a look at my ping monitoring, no issues there it seems (attached)?
Are you all on any of the exchanges on this link: http://help2.talktalk.co.uk/planned-improvements-exchanges (http://help2.talktalk.co.uk/planned-improvements-exchanges)
-
Still having issues here.
Webpages are slow to load with packets being dropped.
-
Jamie I'm not on that list of congested exchanges. I'm still having issues occasionally this morning. I'm sure TT have a problem in their network.
Stuart
-
As soon as I posted on the TalkTalk Community things seemed to improve almost instantly.
It's almost back to normal. For now.
-
Jamie I'm not on that list of congested exchanges. I'm still having issues occasionally this morning. I'm sure TT have a problem in their network.
Stuart
I'd say you're right with a network problem as all the TalkTalk websites are slow to load or fail to load at all for me this morning.
All other sites are fine though
Sent from my Nexus 6 using Tapatalk
-
I'm still having problems. >:(
One site in particular that never finishes loading is none other than forum.kitz.co.uk
I've had quite a struggle in an attempt to make this posting. :(
-
YouTube also doesn't load properly.
Typical that I get a faster connection and I can't use it! >:(
-
YouTube also doesn't load properly.
Typical that I get a faster connection and I can't use it! >:(
I'm still having problems. >:(
One site in particular that never finishes loading is none other than forum.kitz.co.uk
I've had quite a struggle in an attempt to make this posting. :(
I'll put this here in case it was missed in Alec's other thread:
I just had a thought for you both, when you restart the PPPoE session (Dynamic IP DHCP in TalkTalks case), make sure the IP Address changes.
Whenever I manually force the session drop, my IP remains the same on resync.
Keep dropping it until the IP changes (can take a few tries/never happen). Just don't disconnect the modem as B*cat says as a few resyncs of that and you'd be in trouble!
When I had congestion on my exchange, I had to reconnect the session a good few times in an attempt to get a new IP - may solve it?
-
I'm still having problems. >:(
One site in particular that never finishes loading is none other than forum.kitz.co.uk
I've had quite a struggle in an attempt to make this posting. :(
:'(
One of the things that Ive found in the past that most frequently slows things down is loading from an external site. It could perhaps be a google ad, but theres even been cases of a remote hosted avatar which has stopped pages loading properly :(
-
I've dropped the PPPoA session a number of times and now I can use the service - albeit at a somewhat slower throughput speed.
So b*cat is somewhat less grumpy! ;)
-
I don't know how to force a new IP address on the HG612 :(
-
^ I saw that in the other thread, but I couldnt quite work out what you meant. It was unclear what you were using for a router, yet you said the HG612 was doing the routing. :-\
-
HG612 does the routing, router is connected to it and does the WiFi and DHCP server.
-
HG612 does the routing, router is connected to it and does the WiFi and DHCP server.
So the HG612 does the authentication side of things?
Unfortunately no idea how you can do it, maybe purposely "break" the authentication settings on the HG612 then put them back?
Obviously all without losing sync, otherwise leave it half an hour then turn it back on?
-
Yes the HG612 does the authentication side ;)
-
I have noticed today that some websites are taking a long time to load, just like it is losing packets, a reload of the page seems to load it OK but it was not DNS as it had already started accessing the page. This seems to me not unlike the issue we had on TT a while back.
Anyone else seeing this?
Stuart
Whenever I've experienced this behaviour in the past, it's been an MTU issue almost every time :)
-
Whenever I've experienced this behaviour in the past, it's been an MTU issue almost every time :)
Well that does not explain why it has only just started to happen in the past 5-6 days, nor why it is intermittant. Also with lots of people now seeing the issue I doubt everyone has the wrong MTU.
Stuart
-
Still having issues. I literally cannot use the Internet productively when it's in this state.
Think I'll have to tether my 4G phone.
-
Whenever I've experienced this behaviour in the past, it's been an MTU issue almost every time :)
Well that does not explain why it has only just started to happen in the past 5-6 days, nor why it is intermittant. Also with lots of people now seeing the issue I doubt everyone has the wrong MTU.
Stuart
I didn't say it was a client side MTU issue :D
-
I didn't say it was a client side MTU issue :D
Ah .. OK
Stuart
-
I can understand why it might be an MTU issue.
The throughput is still there, it just takes a long time for the connection to react to a command. This is systematic of an MTU issue, so it looks like somebody at TalkTalk has mucked up.
-
Still having issues. I literally cannot use the Internet productively when it's in this state.
Think I'll have to tether my 4G phone.
This doesn't seem to be too widespread though does it? I'm not experiencing it and i'm in the rainy depths of South Wales? ???
Maybe worth calling them and reporting it?
-
Calling TalkTalk - I'd rather not thanks.
-
Calling TalkTalk - I'd rather not thanks.
The TalkTalk Members' Forum (http://www.talktalkmembers.com/) is what I would use . . . if I could be bothered! :D
-
Have your issues been resolved then, bk?
-
Have your issues been resolved then, bk?
No, it still "comes and goes". But knowing that our CP is a bargain-basement entity, I just have occasional bouts of Cat Cursing (and thus wake the spiders who co-reside with me in The Cattery)! ;)
-
Ours is so useless that I'm using my phone tethered to my computer now. Shame the rest of the family can't do the same.
A query has been raised in the TalkTalk Community, regarding the very poor performance of our service.
-
If it is MTU related, then dropping your LAN MTU down may help. I've no idea what might be an appropriate value but 1400 may be a good starting point and then dropping it down another 50 each time you get the issue.
If you were so inclined, you might even use a packet sniffer to see if you're retransmitting lots of stuff to try and figure out the optimium MTU. I would expect it to be temporary, though, at least until someone figures out which bits of the network they broke and sorts it out.
Might also be something entirely unrelated!
-
The TalkTalk Members' Forum (http://www.talktalkmembers.com/) is what I would use . . . if I could be bothered! :D
I already have opened a thread on there however other people keep posting to it so it gets further away from a rep looking at it, and I can no longer lock my threads!
Stuart
-
Calling TalkTalk - I'd rather not thanks.
I found the forums get no response these days, I prefer calling them now. A fault on the forum takes months to resolve with everyone else posting on your thread, pushing you back the queue as broadstairs says.
-
A lot of the OCE replies seem to be telling people to call - but I don't want to call as the phone support is useless! :(
-
A lot of the OCE replies seem to be telling people to call - but I don't want to call as the phone support is useless! :(
Yes Twitter support also now have you call, and the live chat is almost pointless.
-
I don't even get the option to chat anymore :(
Luckily you can always contact the Chief Executive if, things aren't going your way ;)
-
Looks like it was my TP-LINK WDR3600 router issue causing my problems. I've downgraded the firmware and everything is working well once again :)
-
I think we might have a fault.
Upstream SNR has dropped from 6 to 3.8dB at the same sync speed but surely it shouldn't drop this much during the day?
# xdslcmd info --stats
xdslcmd: ADSL driver and PHY status
Status: Showtime
Retrain Reason: 0
Last initialization procedure status: 0
Max: Upstream rate = 10403 Kbps, Downstream rate = 54212 Kbps
Bearer: 0, Upstream rate = 10869 Kbps, Downstream rate = 47465 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 5.8 3.8
Attn(dB): 23.1 0.0
Pwr(dBm): 11.9 5.2
VDSL2 framing
Bearer 0
MSGc: 18 26
B: 51 241
M: 1 1
T: 64 55
R: 12 12
S: 0.0349 0.7080
L: 14688 2870
D: 929 1
I: 64 127
N: 64 254
Counters
Bearer 0
OHF: 47326557 547755
OHFErr: 392 295
RS: 3525550795 1767278
RSCorr: 4412088 4391
RSUnCorr: 14761 0
Bearer 0
HEC: 3424 0
OCD: 102 0
LCD: 102 0
Total Cells: 1083764682 0
Data Cells: 143865658 0
Drop Cells: 0
Bit Errors: 0 0
ES: 118 225
SES: 0 17
UAS: 25 25
AS: 105997
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
PER: 2.23 9.77
OR: 85.72 26.19
AgR: 47550.25 10895.07
Bitswap: 75144/75151 640/644
Total time = 1 days 5 hours 27 min 2 sec
FEC: 4412088 4391
CRC: 392 295
ES: 118 225
SES: 0 17
UAS: 25 25
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 15 minutes time = 12 min 2 sec
FEC: 117206 2
CRC: 3 1
ES: 1 1
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 115784 24
CRC: 0 9
ES: 0 9
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 5 hours 27 min 2 sec
FEC: 1022385 59
CRC: 40 80
ES: 14 76
SES: 0 6
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 3389703 4332
CRC: 352 215
ES: 104 149
SES: 0 11
UAS: 25 25
LOS: 0 0
LOF: 0 0
LOM: 0 0
Since Link time = 1 days 5 hours 26 min 35 sec
FEC: 4412088 4391
CRC: 392 295
ES: 118 225
SES: 0 17
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
-
What would be good is to get some stats monitoring set up to monitor that upstream snr. Its not normal per se, but currently it's not doing anything for a fault to raised.
Its not creating too much of a fuss for ES counts, so I'd say monitoring it would be a good idea.
It could explain your low upload speed. Which if you did resync now would drop further due to the Snr margin.
-
I have issues also. On TalkTalk Fibre here.
The downspeed throughput went down from 40Mb to 11Mb for around an hour until I forced a new connection which seemed to fix things. Ping seems a bit higher than normal. Some websites are taking a bit of time to load (thinkbroadband being one).
until I forced a new connection which seemed to fix things.
May i ask how you did that :-\