Over 55k eth RX errors
Hi everyone, I am running an exit node[1] with nocix (previously known as datashack) and everything worked pretty well until the beginning of March when they moved the physical location of my dedicated server: since then I am registering a lot of eth0 errors[2] and, as atlas is showing, a lot of downtime. Note also that previously I've run the server for like a year without any single eth error! I've already contacted them asking to check the cables and they replied everything it's ok but I am starting to think that *it must be* some kind of hardware issue. I'd like to debug the issue furthermore to excude any software related problem but I don't know what to do, has anybody got a hint? Thanks, patacca [1] https://atlas.torproject.org/#details/2B72D043164D5036BC1087613830E2ED5C6069... [2] http://pastebin.com/7ZYXmTGE
assuming that you run a linux realy, have a look at the interface with ethtool sudo ethtool eth0 and look if it is in full of half duplex mode. If it is in halfduplex mode that will probably be the cause of the errors. you could try setting it to full duplex mode with e. g. 100 mbit/s with the following command: sudo ethtool -s eth0 speed 100 duplex full autoneg off this might crash your network connection, so be prepared for a remote reboot of the machine (or use IPMI if available). On 19.03.2017 11:55, patacca wrote:
Hi everyone, I am running an exit node[1] with nocix (previously known as datashack) and everything worked pretty well until the beginning of March when they moved the physical location of my dedicated server: since then I am registering a lot of eth0 errors[2] and, as atlas is showing, a lot of downtime. Note also that previously I've run the server for like a year without any single eth error! I've already contacted them asking to check the cables and they replied everything it's ok but I am starting to think that *it must be* some kind of hardware issue. I'd like to debug the issue furthermore to excude any software related problem but I don't know what to do, has anybody got a hint?
Thanks, patacca
[1] https://atlas.torproject.org/#details/2B72D043164D5036BC1087613830E2ED5C6069... [2] http://pastebin.com/7ZYXmTGE _______________________________________________ tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Dear patacca, First, thank you for running a relay! Second, great advice from fatal here:
assuming that you run a linux realy, have a look at the interface with ethtool
sudo ethtool eth0
and look if it is in full of half duplex mode. If it is in halfduplex mode that will probably be the cause of the errors. you could try setting it to full duplex mode with e. g. 100 mbit/s with the following command:
sudo ethtool -s eth0 speed 100 duplex full autoneg off
this might crash your network connection, so be prepared for a remote reboot of the machine (or use IPMI if available).
I will add: You are seeing some frame errors, which could point to a bad cable or switch port. I would contact the provider and ask them two things: 1. What is the duplex setting on the switch port? Full? Auto? Half? If they are willing to hard-set it to full, which you then mimic on your end, all the better. 2. Are they seeing errors on input or output on the switch port? If so, ask them to replace the cable. If that doesn't do it, ask them to move you to a new switch port. Good luck! Be well, Rabbi Rob. - -- Rabbi Rob Thomas Team Cymru "It is easy to believe in freedom of speech for those with whom we agree." - Leo McKern -----BEGIN PGP SIGNATURE----- iQIcBAEBCAAGBQJYzpngAAoJEEPoYWL6hfKNgtYP/izB6X6CuRRKmryOKU9jrw++ x3rl+XGVnpoEp8YbFymoMTvUImxImfYhKH6A9P0hIMPce+qZzXkrUxjlKnjgwpJc e7ETA0Cqgjoc7IwbhuKHGXknxOvVeHEdRCXH4NM5RK19nYCrSx5Fw/upOSAoETkk LhfSLRn4KAOQ/GZPiYwGoepMyPoB+vNqgZobQ/1g9mqbdkjjJ2na6pXkCP57J+hk e1Ea2SH18g4GBv29BIXTGGiO66/9wEeF5dpo3IorwexBagNt9ZE9ODvYmzZQxb33 OByQESu7/UsZWDQrEeqhvShDCJ+lOM1MNVON7LwgJgqNR9wFUSgUJbdkGyfMESgO VFjsL0/2qgL1v7cDeDecwitN5Jmyjltt0HaCFGZ34/5Y+DP5nLA1g0t/IScm+57c VwWS0V6+uzhrAgMw7WlCX6K7A13QlcrrEAaJqiuRhMMQDsg7gNs5fshy+9csdSTo vVDH6/8KAtdP1JoSjHPs+0eb3gYlEbTaeoklwDLPDipLF0dL9E+uG2SkYuRkub36 39BIinXlkaCLQsE+e+6Ftk/OEmlXd2QYoVgaoqwWgSsu7FhekQZod/8oRdVkgJFS S3TYwiaaFNv9upqns6N0FjupKo1NqirLWgDaDwWvA9cxXS5rmrS5WoJuEHODqRul MmmEXXTGJisp9WmRhi89 =Bn53 -----END PGP SIGNATURE-----
On 19/03/2017 15:38, fatal wrote:
assuming that you run a linux realy, have a look at the interface with ethtool
Damn it, I knew I would have forgotten something. Yes, the machine is running debian 8.7 root@jah ~ # uname -a Linux jah 3.16.0-4-686-pae #1 SMP Debian 3.16.39-1+deb8u2 (2017-03-07) i686 GNU/Linux
sudo ethtool eth0
and look if it is in full of half duplex mode. If it is in halfduplex mode that will probably be the cause of the errors. you could try setting it to full duplex mode with e. g. 100 mbit/s with the following command:
sudo ethtool -s eth0 speed 100 duplex full autoneg off
this might crash your network connection, so be prepared for a remote reboot of the machine (or use IPMI if available).
Thanks for the advice, it seems to be running in full duplex with auto-negotiation on. I do not know what autonegotiation means, should I try disabling it? Here's the result of ethtool eth0 anyway: root@jah ~ # ethtool eth0 Settings for eth0: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supported pause frame use: No Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Advertised pause frame use: No Advertised auto-negotiation: Yes Speed: 100Mb/s Duplex: Full Port: Twisted Pair PHYAD: 1 Transceiver: internal Auto-negotiation: on MDI-X: on (auto) Supports Wake-on: pumbg Wake-on: g Current message level: 0x00000007 (7) drv probe link Link detected: yes
participants (3)
-
fatal -
patacca -
Rabbi Rob Thomas