Bruce Momjian <bruce(at)momjian(dot)us> writes:
> Tom Lane wrote:
>> Something that might also be interesting is an option to suppress
>> per-command ps_status reporting. On machines where updating ps status
>> takes a kernel call, there's now a pretty good argument why you might
>> want to turn that off and rely on pg_stat_activity instead.
> OK, can I get a timing report from someone with the title on/off that
> shows a difference?
IIRC, newer BSDen use a kernel call for this, so you should be able to
measure it on your own machine. Just tweak ps_status.c to force it to
select PS_USE_NONE instead of PS_USE_SETPROCTITLE to generate a
comparison case. I'll try it on my old HPUX box too.
regards, tom lane
In response to
pgsql-hackers by date
|Next:||From: Josh Berkus||Date: 2006-06-26 17:43:28|
|Subject: Re: Anyone still care about Cygwin? (was Re: [CORE] GPL|
|Previous:||From: Tom Lane||Date: 2006-06-26 17:24:41|
|Subject: pgsql: Change the row constructor syntax (ROW(...)) so that list |
pgsql-patches by date
|Next:||From: Hiroshi Saito||Date: 2006-06-26 17:34:58|
|Subject: Re: MS-VC build patch|
|Previous:||From: Bruce Momjian||Date: 2006-06-26 17:24:39|
|Subject: Re: [HACKERS] Overhead for stats_command_string et al, take 2|