I am using haproxy 1.7 from debian (1.7.5-2~bpo8+1 2017/05/27) as a HTTP front server.
While reading the logs, I noticed that some requests are consistently reported with a high total processing time (around 10 seconds) despite being served very fast to the browser (20-40ms according to firefox network tab) :
[21/Aug/2018:10:33:16.964] HTTP-443~ bck_admin/PRD_20082 0/0/0/16/10000 200 169 - - ---- 67/16/6/7/0 0/0 "GET ...
I am talking about this part specifically :
TR '/' Tw '/' Tc '/' Tr '/' Ta* 0 / 0 / 0 /16 /10000
According to the documentation
Ta (here 10 seconds) is the total processing time in haproxy. Quoting the doc :
- “Ta” is the time the request remained active in haproxy, which is the total time in milliseconds elapsed between the first byte of the request was received and the last byte of response was sent. It covers all possible processing except the handshake (see Th) and idle time (see Ti). There is one exception, if “option logasap” was specified, then the time counting stops at the moment the log is emitted. In this case, a ‘+’ sign is prepended before the value, indicating that the final one will be larger. See “Timers” below for more details.
What could cause the ‘Ta’ to be so high in haproxy while the total duration measured in the browser is only a few dozen of ms?