User comments on ISPs
  >> Entanet International Ltd.


Register (or login) on our website and you will not see this ad.


Pages in this thread: 1 | 2 | (show all)   Print Thread
Standard User legume
(experienced) Wed 09-Jan-13 15:09:11
Print Post

Enta/ukfsn traffic blocking policy.


[link to this post]
 
I notice ukfsn have a revamped website which has this information about traffic management (I assume it comes from Enta but don't know that).

Delivering Quality Service

Our aim is to deliver a positive user experience for all of our customers, primarily for interactive services, for example web browsing, video streaming, VoIP and hosted application access.

In order to deliver a better user experience of interactive services, less priority across the broadband network will be given to non-interactive bulk services (e.g. Peer-to-peer and newsgroups) at times of congestion. However, there are no specific or hard set trigger points at which prioritisation is applied and released on the protocols.

Compared to other communications providers that set specific and permanent maximum limits on speed of access to non-interactive bulk services irrespective of bandwidth availability, these measures are applied dynamically and traffic types are given the highest possible speed according to their priority versus bandwidth availability in real time.


My recent experience on family 30 21cn does not match this description as certain protocols seem to be totally blocked by an actively spoofed RST rather than given lower priority and load on my node does not affect this.

Text only news (message body) is totally blocked for me till 20:00 - this is with enta or ukfsn own servers.

Git pull is a bit more erratic but is often met with a spoofed RST once the transfer starts. I know it is Enta as several times when affected I have tested to multiple servers all fail repeatedly and then at 20::00 all start working perfectly. I don't know whether this blocking is accidental if it's intended then it's not done too well as a clone will work - leading to the pathetic situation of me having to clone 100s of megs worth of git trees just to get the 20k worth of updates I want to test.

I have had a ticket open with Entanet for some time over this but after initial engagement nothing has happened.

I feel anyone reselling Enta should be aware that if they are using Enta's words to describe the product, that they are (currently at least) misleading.
Standard User Uilebheist
(legend) Wed 09-Jan-13 15:34:10
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: legume] [link to this post]
 
Look at this post in the resellers section - they talk aboit vivaciti (via Enta) so that suggetss you'd have the same on ukfsn.
Interestingly I can access ukfsn's news server pretty much 24x7, and it looks like their ADSL customers are blocked by Enta when they try to do so!
(edit: just to clarify I still have a non-ADSL ukfsn account, which gives me access to the news server amongst other things; however I have ADSL from AAISP which of course doesn't suffer from Enta's blocking)

Edited by Uilebheist (Wed 09-Jan-13 15:36:17)

Standard User legume
(experienced) Wed 09-Jan-13 16:33:29
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: Uilebheist] [link to this post]
 
Thanks for the link.

I think I may be joining you in respect of keeping my ukfsn account and getting away from Enta.

I know for ages they have blocked news by protocol and by site (eg. giganews speed test page ) but initially at least it really only was when the node was heavily loaded - so rare, now it's just time based it's become a joke.

The refusal/inability to whitelist their own servers and blocking with spoofed RSTs things like git:// is pathetic and IMO leaves resellers in need of clarifying their advertising to make this clear or pressuring their supplier to mend their blocking policies/config/kit .

I am currently on a rolling month contract - but would well miffed had I just signed up for vdsl with a lock in only to find my ISP was deliberately breaking my connection.


Register (or login) on our website and you will not see this ad.

Standard User Uilebheist
(legend) Wed 09-Jan-13 17:09:58
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: legume] [link to this post]
 
Well, rereading your post now...
Blocking usenet is not good, but blocking git is inexcusable. This would stop me getting my work done!
Standard User HarryHomers
(learned) Wed 09-Jan-13 17:43:07
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: Uilebheist] [link to this post]
 
In reply to a post by Uilebheist:
Look at this post in the resellers section - they talk aboit vivaciti (via Enta) so that suggetss you'd have the same on ukfsn.

I gave my example of what was happening to me in the other thread, I'm on UKFSN through Enta and at the moment I am get news groups at full speed just fine when the node (Bristol) isn't busy.
As far as i can remember haven't the rules on throttling been the same for ages ?
Standard User legume
(experienced) Wed 09-Jan-13 18:08:11
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: HarryHomers] [link to this post]
 
In reply to a post by HarryHomers:
As far as i can remember haven't the rules on throttling been the same for ages ?


For me it's got worse - maybe it's in error or maybe nodes that get more loaded get different policies but on my node Peterborough it seems to be time based for news and random for git://.

Edit - it did used to be just peak and rare for news + never for git.

Currently at 70% while yours is 61% - does this work for you -

http://www.giganews.com/test_connect.html

(must admit I haven't checked for ages) - currently fails for me - hope they haven't changes the link.

Of course another possibility is that I am deemed a bad user in some way - but if this is the case the advertising still needs to be changed. It's the end of my month and I still have 12 gig left - sometimes I am closer than that.

Edited by legume (Wed 09-Jan-13 18:12:02)

Standard User Uilebheist
(legend) Wed 09-Jan-13 18:18:16
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: legume] [link to this post]
 
In reply to a post by legume:
http://www.giganews.com/test_connect.html

(must admit I haven't checked for ages) - currently fails for me - hope they haven't changes the link.

Well, it works from here, it displays two traceroutes from them back to me and it's almost immediate.
If you can't access it, it must be something between Enta and Giganews, or Enta blocking it.
Standard User legume
(experienced) Wed 09-Jan-13 20:06:19
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: Uilebheist] [link to this post]
 
As expected really it fails with a spoofed RST

Edit - forgot to put - my node was at 56% just before 20:00

Text
1
23
45
67
18:20:16.511133 IP 192.168.0.3.34590 > 192.168.0.1.53: 50916+ A? www.giganews.com. (34)
18:20:16.551701 IP 192.168.0.1.53 > 192.168.0.3.34590: 50916 1/2/2 A 216.196.101.133 (126)18:20:16.551837 IP 192.168.0.3.36482 > 216.196.101.133.80: Flags [S], seq 2195509631, win 14600, options [mss 1460,sackOK,TS val 28282220 ecr 0,nop,wscale 7], length 0
18:20:16.645407 IP 216.196.101.133.80 > 192.168.0.3.36482: Flags [S.], seq 3528839124, ack 2195509632, win 14480, options [mss 1380,sackOK,TS val 1246783855 ecr 28282220,nop,wscale 7], length 018:20:16.645430 IP 192.168.0.3.36482 > 216.196.101.133.80: Flags [.], ack 1, win 115, options [nop,nop,TS val 28282314 ecr 1246783855], length 0
18:20:16.645521 IP 192.168.0.3.36482 > 216.196.101.133.80: Flags [P.], seq 1:407, ack 1, win 115, options [nop,nop,TS val 28282314 ecr 1246783855], length 40618:20:16.663830 IP 216.196.101.133.80 > 192.168.0.3.36482: Flags [R.], seq 1:23, ack 407, win 115, length 22


and again just before 20:00

Text
1
23
45
19:56:09.360581 IP 192.168.0.3.36649 > 216.196.101.133.80: Flags [S], seq 621650003, win 14600, options [mss 1460,sackOK,TS val 34035029 ecr 0,nop,wscale 7], length 0
19:56:09.453508 IP 216.196.101.133.80 > 192.168.0.3.36649: Flags [S.], seq 2841200895, ack 621650004, win 14480, options [mss 1380,sackOK,TS val 1252704483 ecr 34035029,nop,wscale 7], length 019:56:09.453531 IP 192.168.0.3.36649 > 216.196.101.133.80: Flags [.], ack 1, win 115, options [nop,nop,TS val 34035122 ecr 1252704483], length 0
19:56:09.453605 IP 192.168.0.3.36649 > 216.196.101.133.80: Flags [P.], seq 1:304, ack 1, win 115, options [nop,nop,TS val 34035122 ecr 1252704483], length 30319:56:09.471008 IP 216.196.101.133.80 > 192.168.0.3.36649: Flags [R.], seq 1:23, ack 304, win 115, length 22


Then works normally post 20:00

Text
1
23
45
67
89
1011
1213
1415
1617
1819
2021
20:00:41.272366 IP 192.168.0.3.53985 > 192.168.0.1.53: 32189+ A? www.giganews.com. (34)
20:00:41.371986 IP 192.168.0.1.53 > 192.168.0.3.53985: 32189 1/2/2 A 216.196.101.133 (126)20:00:41.372108 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [S], seq 1387609103, win 14600, options [mss 1460,sackOK,TS val 34307041 ecr 0,nop,wscale 7], length 0
20:00:41.466800 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [S.], seq 3616911946, ack 1387609104, win 14480, options [mss 1380,sackOK,TS val 1252832100 ecr 34307041,nop,wscale 7], length 020:00:41.466823 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [.], ack 1, win 115, options [nop,nop,TS val 34307135 ecr 1252832100], length 0
20:00:41.466893 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [P.], seq 1:304, ack 1, win 115, options [nop,nop,TS val 34307135 ecr 1252832100], length 30320:00:41.564689 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [.], ack 304, win 122, options [nop,nop,TS val 1252832198 ecr 34307135], length 0
20:00:41.598130 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [P.], seq 1:640, ack 304, win 122, options [nop,nop,TS val 1252832230 ecr 34307135], length 63920:00:41.598140 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [.], ack 640, win 137, options [nop,nop,TS val 34307267 ecr 1252832230], length 0
20:00:41.599132 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [P.], seq 304:628, ack 640, win 137, options [nop,nop,TS val 34307268 ecr 1252832230], length 32420:00:41.735573 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [.], ack 628, win 130, options [nop,nop,TS val 1252832369 ecr 34307268], length 0
20:00:41.945304 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [.], seq 640:2008, ack 628, win 130, options [nop,nop,TS val 1252832576 ecr 34307268], length 136820:00:41.947159 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [.], seq 2008:3376, ack 628, win 130, options [nop,nop,TS val 1252832576 ecr 34307268], length 1368
20:00:41.947169 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [.], ack 3376, win 182, options [nop,nop,TS val 34307616 ecr 1252832576], length 020:00:41.949238 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [P.], seq 3376:4766, ack 628, win 130, options [nop,nop,TS val 1252832576 ecr 34307268], length 1390
20:00:41.949247 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [.], ack 4766, win 205, options [nop,nop,TS val 34307618 ecr 1252832576], length 020:00:41.951125 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [.], seq 4766:6134, ack 628, win 130, options [nop,nop,TS val 1252832576 ecr 34307268], length 1368
20:00:41.953208 IP 216.196.101.133.80 > 192.168.0.3.36691: Flags [.], seq 6134:7502, ack 628, win 130, options [nop,nop,TS val 1252832576 ecr 34307268], length 136820:00:41.953217 IP 192.168.0.3.36691 > 216.196.101.133.80: Flags [.], ack 7502, win 250, options [nop,nop,TS val 34307622 ecr 1252832576], length 0
 SNIP lots more

Edited by legume (Wed 09-Jan-13 20:16:06)

Standard User Uilebheist
(legend) Wed 09-Jan-13 20:35:35
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: legume] [link to this post]
 
Hmm, if you have tcptraceroute (or on Linux "traceroute -T" will do) you can see after which hop you get the RST. If you get the RST from "very near" (after 3 hops for example) then that would help pointing the finger at Enta. From AAISP I get:
traceroute -T www.giganews.com
Password: 
traceroute to www.giganews.com (216.196.101.133), 30 hops max, 60 byte packets
 1  ********  0.295 ms  0.269 ms  0.191 ms
 2  c.gormless.thn.aa.net.uk (90.155.53.53)  21.601 ms  21.924 ms  23.457 ms
 3  c.aimless.thn.aa.net.uk (90.155.53.43)  24.365 ms  25.542 ms  26.200 ms
 4  ge-1-0-0-0.mario.as34270.net (85.91.238.217)  27.762 ms  28.446 ms  29.120 ms
 5  xe-0-0-0-0.yoshi.as34270.net (85.91.232.6)  42.934 ms  43.120 ms  43.063 ms
 6  xe-0-0-1-0.peach.as34270.net (85.91.224.14)  39.229 ms xe-0-0-0-0.peach.as34270.net (85.91.224.26)  38.683 ms xe-0-0-1-0.peach.as34270.net (85.91.224.14)  38.930 ms
 7  * * *
 8  te0-1-0-4.ccr21.lpl01.atlas.cogentco.com (130.117.51.254)  30.940 ms  29.956 ms te0-3-0-4.ccr21.lpl01.atlas.cogentco.com (154.54.62.145)  30.648 ms
 9  te0-6-0-4.ccr21.bos01.atlas.cogentco.com (154.54.85.209)  97.618 ms te0-3-0-3.ccr21.bos01.atlas.cogentco.com (154.54.31.229)  95.482 ms te0-5-0-4.ccr21.bos01.atlas.cogentco.com (154.54.85.213)  96.467 ms
10  te0-0-0-18.mpd21.jfk02.atlas.cogentco.com (154.54.44.5)  103.554 ms te0-7-0-11.mpd21.jfk02.atlas.cogentco.com (154.54.6.2)  105.874 ms te0-0-0-11.mpd21.jfk02.atlas.cogentco.com (154.54.44.17)  104.453 ms
11  te0-3-0-3.mpd21.dca01.atlas.cogentco.com (154.54.1.105)  110.826 ms te0-3-0-6.mpd21.dca01.atlas.cogentco.com (154.54.41.21)  107.468 ms  108.957 ms
12  te0-3-0-7.mpd21.iad02.atlas.cogentco.com (154.54.1.230)  109.220 ms te0-0-0-2.mpd21.iad02.atlas.cogentco.com (154.54.1.206)  109.526 ms te0-3-0-5.mpd21.iad02.atlas.cogentco.com (154.54.41.246)  110.347 ms
13  38.122.62.202 (38.122.62.202)  111.188 ms 38.122.62.210 (38.122.62.210)  111.546 ms 38.122.62.202 (38.122.62.202)  111.929 ms
14  g3-16.gw2.dca.giganews.com (216.196.96.61)  113.564 ms  114.208 ms  115.473 ms
15  216.196.101.133 (216.196.101.133)  117.606 ms  118.000 ms  108.502 ms
Standard User Uilebheist
(legend) Wed 09-Jan-13 20:37:47
Print Post

Re: Enta/ukfsn traffic blocking policy.


[re: Uilebheist] [link to this post]
 
And for comparison, to news.ukfsn.org (port 119)
traceroute -T -p 119 news.ukfsn.org
traceroute to news.ukfsn.org (77.75.108.119), 30 hops max, 60 byte packets
 1  ***********  0.234 ms  0.222 ms  0.264 ms
 2  c.gormless.thn.aa.net.uk (90.155.53.53)  21.188 ms  21.789 ms  22.952 ms
 3  d.aimless.thn.aa.net.uk (90.155.53.44)  24.097 ms  25.011 ms  26.179 ms
 4  195.50.95.65 (195.50.95.65)  151.970 ms  152.165 ms  152.500 ms
 5  GOSCOMB-TEC.car5.London1.Level3.net (212.187.200.102)  29.685 ms  30.608 ms  31.301 ms
 6  xe-0-0-0.rt0.the.uk.goscomb.net (93.89.91.13)  32.727 ms  31.929 ms  33.188 ms
 7  xe-0-0-1-0.rt0.sov.uk.goscomb.net (46.17.60.34)  33.402 ms  21.855 ms  22.716 ms
 8  ae0.cs0.sov.uk.goscomb.net (46.17.60.170)  27.330 ms  22.966 ms  23.325 ms
 9  xe-0-1-1.cs0.mer.uk.goscomb.net (93.89.90.146)  23.666 ms  23.988 ms  24.807 ms
10  news.ukfsn.org (77.75.108.119)  25.897 ms  26.721 ms  27.078 ms
Pages in this thread: 1 | 2 | (show all)   Print Thread

Jump to