Project

General

Profile

tickets #20440

mirrorbrain is extremely slow

Added by mchandras@suse.de almost 5 years ago. Updated almost 5 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
Start date:
Due date:
% Done:

0%

Estimated time:

Description

Hi,

Since this morning I am having several issues with the openSUSE mirrors
on both TW and Leap repositories. I believe that problem is with
mirrorbrain since as far as I can a direction happens on every rpm
download and that redirection seems to take several seconds.

For example, here is some timing results from an openSUSE Leap 42.2 host

$ curl --trace-time -v -L
http://download.opensuse.org/distribution/leap/42.2/repo/oss/suse/x86_64/libreoffice-5.1.5.2-12.1.x86_64.rpm

/dev/null
[...]
14:46:28.213196 * TCP_NODELAY set
[...]
14:46:44.902111 * Connected to mirror.ox.ac.uk (163.1.160.163) port 80 (#1)

so it took 16 seconds to send me to a proper mirror

Sometimes it takes even longer without success

$ curl --trace-time -v -L
http://download.opensuse.org/distribution/leap/42.2/repo/oss/suse/x86_64/libreoffice-5.1.5.2-12.1.x86_64.rpm

/dev/null
[...]
14:48:31.269343 * TCP_NODELAY set
[...]
0 0 0 0 0 0 0 0 --:--:-- 0:02:09 --:--:--
014:50:40.852188 * connect to 2001:67c:2178:8::13 port 80 failed:
Connection timed out
14:50:40.856114 * connect to 195.135.221.134 port 80 failed: Connection
timed out
14:50:40.856185 * Failed to connect to download.opensuse.org port 80:
Connection timed out
14:50:40.856223 * Closing connection 0
curl: (7) Failed to connect to download.opensuse.org port 80: Connection
timed out

The results are pretty similar in the TW repositories

Seems like there was a complain about mirrorbrain yesterday as well

https://lists.opensuse.org/opensuse-factory/2017-07/msg00242.html

I am not sure if you are already aware of it so I thought to let you know

--
markos

SUSE LINUX GmbH | GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg) Maxfeldstr. 5, D-90409, Nürnberg

History

#1 Updated by maxmaher almost 5 years ago

[3634793.285047] TCP: Possible SYN flooding on port 80. Sending cookies.
[3634793.285176] TCP: Possible SYN flooding on port 80. Sending cookies.
[3634798.313257] net_ratelimit: 596 callbacks suppressed
[3634798.313262] TCP: Possible SYN flooding on port 80. Sending cookies.
[3634798.313340] TCP: Possible SYN flooding on port 80. Sending cookies.
[3634798.313349] TCP: Possible SYN flooding on port 80. Sending cookies.

thats what i see on pontifex3 all the time...

something is really not ok, but i don't have a clue atm

#2 Updated by pjessen almost 5 years ago

  • Private changed from Yes to No

FWIW, it's working fine atm.

#3 Updated by markoschandras almost 5 years ago

pjessen wrote:

FWIW, it's working fine atm.

Hmm not for me. Both repos are still terribly slow for me.

#4 Updated by lnussel almost 5 years ago

for some reason, maybe related to the network changes this morning mirrorbrain forgot the content of all TW mirrors. I've issued a manual scan, now we have some back at least. This should free some bandwidth again.

#5 Updated by lnussel almost 5 years ago

/var/log ran out of space. I've deleted March and April logs for download.o.o, they are archived at langley. 21G free again, I hope that helps until tomorrow

#6 Updated by markoschandras almost 5 years ago

lnussel wrote:

/var/log ran out of space. I've deleted March and April logs for download.o.o, they are archived at langley. 21G free again, I hope that helps until tomorrow

Ah great thanks. Both repos seem to behave normally again

#7 Updated by maxmaher almost 5 years ago

  • Status changed from New to Resolved

Yes Ludwig found the answer to our problem,

indeed it was related to the NT outage.

Also available in: Atom PDF