Eelco Chaudron <[email protected]> writes:

> This patch adds a coverage counter for long poll interval events
> which might help debugging the issue.
>
> Signed-off-by: Eelco Chaudron <[email protected]>
> ---

I prefer the commit log from:

  https://www.mail-archive.com/[email protected]/msg79323.html

There's a bit more justification on why we should expose via the
coverage counters.  For instance, this data is currently available via
the journal / system log.  But it isn't available programmatically or if
the journal entry rotates.

>  lib/timeval.c |    4 ++++
>  1 file changed, 4 insertions(+)
>
> diff --git a/lib/timeval.c b/lib/timeval.c
> index 193c7bab1..0abe7e555 100644
> --- a/lib/timeval.c
> +++ b/lib/timeval.c
> @@ -41,6 +41,8 @@
>  
>  VLOG_DEFINE_THIS_MODULE(timeval);
>  
> +COVERAGE_DEFINE(long_poll_interval);
> +
>  #if !defined(HAVE_CLOCK_GETTIME)
>  typedef unsigned int clockid_t;
>  static int clock_gettime(clock_t id, struct timespec *ts);
> @@ -644,6 +646,8 @@ log_poll_interval(long long int last_wakeup)
>          const struct rusage *last_rusage = get_recent_rusage();
>          struct rusage rusage;
>  
> +        COVERAGE_INC(long_poll_interval);
> +
>          if (!getrusage_thread(&rusage)) {
>              VLOG_WARN("Unreasonably long %lldms poll interval"
>                        " (%lldms user, %lldms system)",
>
> _______________________________________________
> dev mailing list
> [email protected]
> https://mail.openvswitch.org/mailman/listinfo/ovs-dev

_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to