Hello Serge, this list is english-speaking, but please see my responses below.
On Thu, Feb 04, 2010 at 02:52:11PM +0100, Reynier Serge wrote: > Bonjour, > > j'ai mis un haproxy en place pour des connexions MySQL et j'aimerai > loguer seulement les erreurs et warning de haproxy visible sur la page > de stats. Le mode de log est donc tcplog, voici un extrait de mon > fichier de conf : > > listen MySQL x.x.x.x:3306 > mode tcp > option tcplog > option dontlog-normal > no option log-separate-errors > log 127.0.0.1 local0 debug > # option tcpka > maxconn 10000 > timeout connect 3s > timeout client 3s > timeout server 3s > server SRV1 x.x.x.x:3306 check inter 10s rise 3 fall 3 weight 1 > server SRV2 x.x.x.x:3306 check inter 10s rise 3 fall 3 weight 1 > > Quand je met le niveau de log a debug on a beaucoup trop de ligne avec > des flags "cD", si je met le niveau a notice plus rien ... Traffic logs are emitted at the "info" level. "notice" only reports more important events such as servers going UP/DOWN. The fact that you get lots of "cD" in your logs is because neither the client nor the server has anything to say within your 3s timeout, most likely because you're using persistent connections and don't have enough traffic to ensure that there will always be some requests within 3s. As Mariusz suggested it, you should really raise those values. Also, using 3s is really not good, because it is the default TCP retransmit delay. That means that if one packet is lost over the network, it will not have enough time to be retransmitted and the timeout will strike first. I generally suggest 4-5 seconds for "timeout connect", and in your case, you should have way larger ones for "timeout client" and "timeout server" (eg: 1m ?). > En fait je voudrais simplement capturer les quelques warning et error > que j'ai sur la page de stats. They're all there in your logs. "cD" is considered as an abnormal session termination since there was a timeout. It is polluting your logs due to your low timeouts, but it is very likely that the other errors you see are precisely caused by aborted sessions consecutively to lost packets. Hoping this helps, Willy
