You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@trafficserver.apache.org by Jean Baptiste Favre <we...@jbfavre.org> on 2013/12/18 12:53:49 UTC
Monitoring question
Hello,
I'm trying to monitor y trafficserver and especially average transaction
response time.
As a start, I had a look at tstop source code, which use
proxy.process.http.total_transactions_time and
proxy.process.http.incoming_requests from stats_over_http plugin.
It builds client_req_time, dividing total_transactions_time by
incoming_requests which give an average time per request.
So far, so good.
But I also see in tools/tstop/stats.h is divided by 10.000.000 to get
milliseconds (line 286 & 292).
I could understand 1.000.000 to convert nanoseconds into millisecond;
I also could understand 1.000 to convert microseconds into millisecond
or nanoseconds into microsecond. But here, why using 10.000.000 ?
Since my monitoring will mimic tstop, I want to make sure I have the
right units :)
Regards,
Jean-Baptiste