Error message via nntp:Too Many Connections [FORUMS-READERS allows 50 per host/0 per feed (0 == noli

martin_helm · October 24, 2012, 1:42pm

I get since yesterday night the error message

Too Many Connections [FORUMS-READERS allows 50 per host/0 per feed (0 == nolimit)] (Typhoon v2.1.2.394)

and cannot see a valid reason, there does not seem to be an excessive amount of connections

martinh@ganymed:~> sudo netstat -tup | grep opensuse
tcp        0      0 ganymed.fritz.box:44522 forums.opensuse.or:nntp CLOSE_WAIT  7523/thunderbird-bi 
martinh@ganymed:~>

Anyone else who sees something like that or has an explanation.
I use Thunderbird 16 as reader on openSUSE 12.2 64 bit KDE 4.9.
Restart of TB or even the machine does not help (I wanted to be sure there is no connection so I even restarted the notebook). No other machine is connected

martin_helm · October 24, 2012, 1:59pm

The problem suddenly disappeared and solved itself, no idea why.

–
PC: oS 12.2 x86_64 | i7-2600@3.40GHz | 16GB | KDE 4.8.5 | GeForce GT 420
ThinkPad E320: oS 12.2 x86_64 | i3@2.30GHz | 8GB | KDE 4.9.2 | HD 3000
eCAFE 800: oS 11.4 i586 | AMD Geode LX 800@500MHz | 512MB | lamp server

hcvv · October 24, 2012, 2:58pm

But it is allways useful to report such things when they happen. Thanks.

caf4926 · October 24, 2012, 4:28pm

I can’t tell you why. But I have had issues of time outs.
It’s still not quite right IMO, though fixes have been applied.
And fetched messages are taking ages to display in TBird

I’m working with @hendersj in admin on this

Will let you know …

djh-novell · October 24, 2012, 4:44pm

caf4926 wrote:
> martin_helm;2498430 Wrote:
>> The problem suddenly disappeared and solved itself, no idea why.
>>
>> –
>> PC: oS 12.2 x86_64 | i7-2600@3.40GHz | 16GB | KDE 4.8.5 | GeForce GT
>> 420
>> ThinkPad E320: oS 12.2 x86_64 | i3@2.30GHz | 8GB | KDE 4.9.2 | HD 3000
>> eCAFE 800: oS 11.4 i586 | AMD Geode LX 800@500MHz | 512MB | lamp server
>
> I can’t tell you why. But I have had issues of time outs.
> It’s still not quite right IMO, though fixes have been applied.
> And fetched messages are taking ages to display in TBird
>
> I’m working with @hendersj in admin on this
>
> Will let you know …

FWIW, I haven’t seen any errors but today I am sometimes seeing it take
a long time to establish a connection to update NNTP headers. Most of
the time it’s fine but occasionally it hangs very noticeably.

martin_helm · October 24, 2012, 4:53pm

Am 24.10.2012 16:36, schrieb caf4926:
> I can’t tell you why. But I have had issues of time outs.
> It’s still not quite right IMO, though fixes have been applied.
> And fetched messages are taking ages to display in TBird
>
> I’m working with @hendersj in admin on this
>
> Will let you know …
>
Thanks for the feedback the problem came again 10-15 minutes ago and
went away maybe 2 minutes ago just to let you know.
It not so bad that it is unusable, just a glitch.

–
PC: oS 12.2 x86_64 | i7-2600@3.40GHz | 16GB | KDE 4.8.5 | GeForce GT 420
ThinkPad E320: oS 12.2 x86_64 | i3@2.30GHz | 8GB | KDE 4.9.2 | HD 3000
eCAFE 800: oS 11.4 i586 | AMD Geode LX 800@500MHz | 512MB | lamp server

hendersj · October 24, 2012, 7:59pm

On Wed, 24 Oct 2012 11:46:01 +0000, martin helm wrote:

> Anyone else who sees something like that or has an explanation.

The new appliance that went in (some here were testing it last week) went
live yesterday, and it seems that there was some tuning needed to
increase the number of concurrent connections allowed on the NNTP port.

Should be fixed now.

Jim

–
Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

robin_listas · October 25, 2012, 4:48am

On 2012-10-24 13:46, martin helm wrote:

> Anyone else who sees something like that or has an explanation.
> I use Thunderbird 16 as reader on openSUSE 12.2 64 bit KDE 4.9.

Yes, I saw it yesterday when switching off the machine in tty10, as warnings from leafnode.

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” (Minas Tirith))

martin_helm · October 25, 2012, 1:44pm

Am 24.10.2012 19:59, schrieb Jim Henderson:
> On Wed, 24 Oct 2012 11:46:01 +0000, martin helm wrote:
>
>> Anyone else who sees something like that or has an explanation.
>
> The new appliance that went in (some here were testing it last week) went
> live yesterday, and it seems that there was some tuning needed to
> increase the number of concurrent connections allowed on the NNTP port.
>
> Should be fixed now.
>
> Jim
>
>
>
Looks good, the problem did not come back. Thanks to everyone involved.

–
PC: oS 12.2 x86_64 | i7-2600@3.40GHz | 16GB | KDE 4.8.5 | GeForce GT 420
ThinkPad E320: oS 12.2 x86_64 | i3@2.30GHz | 8GB | KDE 4.9.2 | HD 3000
eCAFE 800: oS 11.4 i586 | AMD Geode LX 800@500MHz | 512MB | lamp server

djh-novell · October 25, 2012, 3:19pm

Martin Helm wrote:
> Looks good, the problem did not come back. Thanks to everyone involved.

I just experienced the slow NNTP access problem again. This time I
noticed that the status line was showing ‘Trying to connect to
forums.opensuse.org’ (or something very much like that) so I tried ping.
The results were:

$ ping forums.opensuse.org
ping: unknown host forums.opensuse.org
$ ping forums.opensuse.org
PING forums.opensuse.org (130.57.66.6) 56(84) bytes of data.
64 bytes from forums.opensuse.org (130.57.66.6): icmp_seq=1 ttl=47
time=243 ms
64 bytes from forums.opensuse.org (130.57.66.6): icmp_seq=2 ttl=47
time=224 ms
64 bytes from forums.opensuse.org (130.57.66.6): icmp_seq=3 ttl=47
time=246 ms

— forums.opensuse.org ping statistics —
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 224.840/238.206/246.171/9.517 ms

I was able to reach other web sites whilst the first ping was running,
so I don’t think the problem is at our end.

hendersj · October 25, 2012, 6:41pm

On Thu, 25 Oct 2012 13:19:13 +0000, Dave Howorth wrote:

> I just experienced the slow NNTP access problem again. This time I
> noticed that the status line was showing ‘Trying to connect to
> forums.opensuse.org’ (or something very much like that) so I tried ping.

If you see it again, a traceroute would be helpful.

Thanks,

Jim

–
Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

robin_listas · October 25, 2012, 9:59pm

On 2012-10-25 18:41, Jim Henderson wrote:
> On Thu, 25 Oct 2012 13:19:13 +0000, Dave Howorth wrote:
>
>> I just experienced the slow NNTP access problem again. This time I
>> noticed that the status line was showing ‘Trying to connect to
>> forums.opensuse.org’ (or something very much like that) so I tried ping.
>
> If you see it again, a traceroute would be helpful.

What I see is very slow nntp right now, using leaf node. Below 200b/s. Had to abort.

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” (Minas Tirith))

hendersj · October 25, 2012, 10:29pm

On Thu, 25 Oct 2012 19:59:40 +0000, Carlos E. R. wrote:

>> If you see it again, a traceroute would be helpful.
>
> What I see is very slow nntp right now, using leaf node. Below 200b/s.
> Had to abort.

What does traceroute say your route to the server is?

Jim

–
Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

robin_listas · October 26, 2012, 1:29am

On 2012-10-25 22:29, Jim Henderson wrote:
> On Thu, 25 Oct 2012 19:59:40 +0000, Carlos E. R. wrote:

>> What I see is very slow nntp right now, using leaf node. Below 200b/s.
>> Had to abort.
>
> What does traceroute say your route to the server is?

It didn’t occur to me to try, sorry. Instead I forced a new connection (I’m on mobile internet now)
and it worked a bit better. And now it is working much better, peaks at 7 KB, which is about what
I’m used to see (dunno why not at 100 KB, though).

I don’t know if it is a local or remote issue. I changed the config to try both novell and opensuse
servers, no appreciable difference.

In any case, the current route is:


> minas-tirith:~ # traceroute -T -p 119 forums.opensuse.org
> traceroute to forums.opensuse.org (130.57.66.6), 30 hops max, 40 byte packets using TCP
>  1  * * *
>  2  * * *
>  3  * * *
>  4  * * *
>  5  bcn-b1-link.telia.net (213.155.129.249)  3140.519 ms   3139.386 ms bcn-b1-link.telia.net (213.155.129.253)  3158.230 ms
>  6  prs-bb2-link.telia.net (80.91.253.99)  4118.144 ms prs-bb2-link.telia.net (80.91.247.14)  4116.998 ms prs-bb1-link.telia.net (80.91.253.252)  3154.840 ms
>  7  ash-bb1-link.telia.net (80.91.254.214)  148.581 ms ash-bb1-link.telia.net (80.91.252.36)  158.400 ms ash-bb1-link.telia.net (80.91.251.247)  177.348 ms
>  8  TenGigE0-2-0-0.GW1.IAD8.ALTER.NET (63.125.125.41)  176.197 ms   165.056 ms   274.626 ms
>  9  0.xe-5-0-0.XL4.IAD8.ALTER.NET (152.63.36.50)  162.794 ms   171.710 ms   170.580 ms
> 10  0.ge-4-1-0.XT2.DEN4.ALTER.NET (152.63.114.201)  270.092 ms   268.856 ms   267.633 ms
> 11  POS7-0.GW3.DEN4.ALTER.NET (152.63.72.73)  266.448 ms   265.286 ms   264.146 ms
> 12  unknown.customer.alter.net (65.206.183.22)  263.009 ms   261.851 ms   260.702 ms
> 13  192.94.118.247 (192.94.118.247)  219.464 ms   218.752 ms   217.576 ms
> 14  * * *
> 15  forums.opensuse.org (130.57.66.6)

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” (Minas Tirith))

djh-novell · October 26, 2012, 3:22pm

Jim Henderson wrote:
> What does traceroute say your route to the server is?

I just had a timeout whilst posting. traceroute says:


# traceroute forums.opensuse.org
traceroute to forums.opensuse.org (130.57.66.6), 30 hops max, 40 byte
packets
1  fw.lmb.internal (10.0.0.1)  0.224 ms   0.102 ms   0.150 ms
2  gw-741.route-hosp.net.cam.ac.uk (131.111.85.254)  0.466 ms   0.417
ms   0.402 ms
3  route-hosp.route-mill.net.cam.ac.uk (192.84.5.109)  0.493 ms   0.445
ms   0.436 ms
4  route-mill.route-enet.net.cam.ac.uk (192.84.5.94)  0.623 ms   0.522
ms   0.528 ms
5  xe-11-3-0.camb-rbr1.eastern.ja.net (146.97.130.1)  0.424 ms   0.468
ms   0.475 ms
6  xe-2-0-0.lond-rbr1.eastern.ja.net (146.97.65.33)  2.169 ms   2.185
ms   2.116 ms
7  ae3.lond-sbr4.ja.net (146.97.35.125)  2.242 ms   2.197 ms   2.146 ms
8  ldn-b4-link.telia.net (213.248.84.177)  2.227 ms   2.302 ms   2.171 ms
9  ldn-bb2-link.telia.net (80.91.249.181)  2.339 ms
ldn-bb2-link.telia.net (80.91.246.146)  2.284 ms ldn-bb1-link.telia.net
(80.91.246.96)  73.561 ms
10  ash-bb1-link.telia.net (80.91.246.66)  77.413 ms
ash-bb1-link.telia.net (213.248.65.98)  150.896 ms
ash-bb1-link.telia.net (80.91.246.68)  85.756 ms
11  TenGigE0-2-0-0.GW1.IAD8.ALTER.NET (63.125.125.41)  81.136 ms
GigabitEthernet2-0-0.GW8.IAD8.ALTER.NET (63.65.76.189)  79.116 ms
80.860 ms
12  0.xe-4-0-1.XL3.IAD8.ALTER.NET (152.63.33.42)  81.384 ms   81.365 ms
81.838 ms
13  0.ge-1-1-0.XT1.DEN4.ALTER.NET (152.63.88.54)  236.769 ms   236.640
ms   236.631 ms
14  POS6-0.GW3.DEN4.ALTER.NET (152.63.72.69)  236.413 ms   236.500 ms
236.532 ms
15  unknown.customer.alter.net (65.206.183.22)  244.281 ms   244.315 ms
244.244 ms
16  192.94.118.247  244.194 ms   244.527 ms   243.960 ms
17  forums.opensuse.org (130.57.66.6)  245.138 ms   245.035 ms   257.812 ms
cpepc210-3:~ # traceroute forums.opensuse.org
traceroute to forums.opensuse.org (130.57.66.6), 30 hops max, 40 byte
packets
1  fw.lmb.internal (10.0.0.1)  0.192 ms   0.189 ms   0.180 ms
2  gw-741.route-hosp.net.cam.ac.uk (131.111.85.254)  0.433 ms   0.301
ms   0.302 ms
3  route-hosp.route-mill.net.cam.ac.uk (192.84.5.109)  0.457 ms   0.430
ms   0.451 ms
4  route-mill.route-enet.net.cam.ac.uk (192.84.5.94)  0.536 ms   0.548
ms   0.535 ms
5  xe-11-3-0.camb-rbr1.eastern.ja.net (146.97.130.1)  0.436 ms   0.531
ms   0.484 ms
6  xe-2-0-0.lond-rbr1.eastern.ja.net (146.97.65.33)  2.142 ms   2.113
ms   2.041 ms
7  ae3.lond-sbr4.ja.net (146.97.35.125)  2.182 ms   2.148 ms   2.185 ms
8  ldn-b4-link.telia.net (213.248.84.177)  2.124 ms   2.160 ms   2.156 ms
9  ldn-bb2-link.telia.net (80.91.249.181)  2.216 ms
ldn-bb2-link.telia.net (80.91.246.146)  2.210 ms ldn-bb1-link.telia.net
(80.91.246.96)  2.180 ms
10  ash-bb1-link.telia.net (80.91.246.66)  77.144 ms
ash-bb1-link.telia.net (213.248.65.98)  108.215 ms
ash-bb1-link.telia.net (80.91.246.68)  77.459 ms
11  TenGigE0-2-0-0.GW1.IAD8.ALTER.NET (63.125.125.41)  83.035 ms
GigabitEthernet2-0-0.GW8.IAD8.ALTER.NET (63.65.76.189)  81.302 ms
83.113 ms
12  0.xe-4-0-1.XL3.IAD8.ALTER.NET (152.63.33.42)  81.312 ms   81.377 ms
81.081 ms
13  0.ge-1-1-0.XT1.DEN4.ALTER.NET (152.63.88.54)  226.450 ms   224.488
ms   224.472 ms
14  POS6-0.GW3.DEN4.ALTER.NET (152.63.72.69)  224.587 ms   224.341 ms
224.458 ms
15  unknown.customer.alter.net (65.206.183.22)  232.706 ms   232.573 ms
232.594 ms
16  192.94.118.247  238.903 ms   238.488 ms   238.859 ms
17  forums.opensuse.org (130.57.66.6)  240.362 ms   239.881 ms   239.816 ms
#

MatthewEhle · October 26, 2012, 6:10pm

It looked like we might have been hitting even the new limit. I bumped it up again. How does it look now?

hendersj · October 26, 2012, 6:21pm

On Thu, 25 Oct 2012 23:29:00 +0000, Carlos E. R. wrote:

> It didn’t occur to me to try, sorry. Instead I forced a new connection
> (I’m on mobile internet now)
> and it worked a bit better. And now it is working much better, peaks at
> 7 KB, which is about what I’m used to see (dunno why not at 100 KB,
> though).
>
> I don’t know if it is a local or remote issue. I changed the config to
> try both novell and opensuse servers, no appreciable difference.

Looking at your traceroute, it looks like you’re getting hammered by
telia. Those rtt responses are multiple seconds from each of those nodes.

Jim

–
Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

hendersj · October 26, 2012, 6:22pm

On Fri, 26 Oct 2012 13:22:12 +0000, Dave Howorth wrote:

> I just had a timeout whilst posting. traceroute says:

Looks like Matt’s made some adjustments, any better now?

I see that you’re going through the telia.net connections that Carlos is
seeing poor response times from, though yours aren’t as bad.

Jim

–
Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

robin_listas · October 26, 2012, 7:35pm

On 2012-10-26 18:21, Jim Henderson wrote:
> On Thu, 25 Oct 2012 23:29:00 +0000, Carlos E. R. wrote:

> Looking at your traceroute, it looks like you’re getting hammered by
> telia. Those rtt responses are multiple seconds from each of those nodes.

But that traceroute is when it is working right…

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” (Minas Tirith))

robin_listas · October 26, 2012, 7:35pm

On 2012-10-26 18:22, Jim Henderson wrote:

> I see that you’re going through the telia.net connections that Carlos is
> seeing poor response times from, though yours aren’t as bad.

Notice that my trace uses port 119 tcp, not the default setting.

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” (Minas Tirith))