On 9 Nov. 2016, at 09:00, Tobias Sachs tobias@tobiassachs.de wrote:
Hey,
my Tor relay https://atlas.torproject.org/#details/B567E8E39641F61091C1F2CAAAF73D3D1BF9CF... balks with the following message which is repeated in the log with the "info" logging level up to 23251 times at the exact same millisecond.
I assume you are talking about the router_pick_published_address message.
The relay responds at :9030/tor/server/authority with a 503 error.
This is a temporary error when your relay is overloaded.
I just tried http://167.114.245.102:9030/tor/server/authority and it delivered the descriptor fine, with no error.
I logged a ticket:
Relays should log a message when they return a 503 error to a client https://trac.torproject.org/projects/tor/ticket/20611
router_pick_published_address(): Success: chose address 'x.x.x.x'
This message is not an error, it is a success.
But I can see how it would be annoying:
Rate limit router_pick_published_address log message https://trac.torproject.org/projects/tor/ticket/20610
This is a counting from the log with the corresponding timestamp.
Nov 01 18:37:13.000 - 23251 Nov 01 18:37:22.000 - 59 Nov 01 18:38:17.000 - 23195 Nov 01 18:38:22.000 - 57 Nov 01 19:19:22.000 - 50 Nov 01 19:19:23.000 - 312 Nov 01 19:19:24.000 - 333 Nov 01 19:19:25.000 - 74 Nov 01 19:20:22.000 - 52 Nov 01 19:38:17.000 - 23207 Nov 01 20:38:17.000 - 23232
140 times exactly 49 repeats in around 2 hours.
Interesting is that there is no error about this 503 error in the log only my monitoring is aware of the issue.
I hope you coul'd help me out with this issue.
Maybe you should turn your log level down to notice? The less you log, the better.
T