package: jenkins.debian.org
(this has been ongoing for month already)
< h01ger> mapreri: i think we need to put restarting network into some sort of
'cronjob' (probably only if network is down), ionos5 is been gone since several
hours and thus half the amd64 builders are down now
< h01ger> not sure how to test if network is up, ping -c 1 8.8.8.8 ?
< mapreri> fwiw, testing the network is actually this very easy:
< mapreri> mattia@ionos5-amd64 ~ % ip -br a
< mapreri> lo UNKNOWN127.0.0.1/8 ::1/128
< mapreri> eth0 UP fe80::1:5dff:fedf:3b89/64
< mapreri> mattia@ionos5-amd64 ~ % sudo service networking restart
< mapreri> mattia@ionos5-amd64 ~ % ip -br a
< mapreri> lo UNKNOWN127.0.0.1/8 ::1/128
< mapreri> eth0 UP 85.184.249.130/32
fe80::1:5dff:fedf:3b89/64
< mapreri> mattia@ionos5-amd64 ~ %
< mapreri>
< mapreri> but still, I'd really really prefer to figure out what's happening
inside those 4 affected hosts :(
< h01ger> what are the 4 nodes?
< h01ger> all ionos/ffm datacenter?
< h01ger> (ffm=frankfurt/main)
< mapreri> yes
< mapreri> 5 6 15 16
* | h01ger will file a bug using this irc log
< mapreri> today it seems to be only 5, for whatever reason.
< h01ger> i think i might have seen 2+12 loosing network too, but that might
have been other i386 related failures. i dont think i've seen this with 1+11
< mapreri> (fwiw, `ip -br -4 a show dev eth0` is currently the shortest command
I use to get a parsable ip4 address)
< h01ger> s#i dont think i've seen this with 1+11#i haven't seen this with 1+11#
< h01ger> those addresses are assigned by dhcp, arent they?
< mapreri> they are
< h01ger> hmm
< mapreri> but it doesn't happen with 3 7 10 that are also there
< mapreri> so I don't think we can blame ionos
< h01ger> in ffm you mean
< mapreri> yep
< h01ger> nods
--
cheers,
Holger
⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ holger@(debian|reproducible-builds|layer-acht).org
⢿⡄⠘⠷⠚⠋⠀ OpenPGP: B8BF54137B09D35CF026FE9D 091AB856069AAA1C
⠈⠳⣄
Manchmal kommt der Wind von Lee. (Konny)
signature.asc
Description: PGP signature
