responding for historical purposes.

i recompiled with ares2.. time skew is no longer an issue.

On Fri, Jan 30, 2015 at 9:49 AM, Jason Strongman
<jasonstrongman2016_at_gmail.com> wrote:
> are there any known issues with the curl -w time measurements?
>
> i noticed that the curl measurements are a bit higher than whats seen
> in a single transaction's corresponding system trace output(strace) or
> packet capture(tcpdump).
>
> for example, in one of my tests, curl reports name lookup as 28msec.
>
> time_namelookup 0.028
>
> when looking at the system trace output and packet capture, the name
> lookup from start to finish only took 18msec. where is curl getting
> the additional 10msec?
> this occurs on all my tests. curl time_namelookup values are much
> higher than whats seen in a pcap/system-trace-output.
> actually all curl time measurements are much higher than whats seen in
> a corresponding pcap.
>
> please advise
>
> thanks!
-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-users
FAQ: http://curl.haxx.se/docs/faq.html
Etiquette: http://curl.haxx.se/mail/etiquette.html
Received on 2015-01-30