If there is something out there that does for perl memory usage analysis
what NYTProf does for profiling, then I am >dying< to get my grubby
little hands on it.
My ideal tool would give me a breakdown of the memory used by current
SVs by package. Not sure if I'm saying that right or if this i
B::TerseSize might be what you want, and Apache2 has a Status handler
to enable that.
http://search.cpan.org/dist/mod_perl/docs/api/Apache2/Status.pod#StatusTerseSizeMainSummary
In a standalone script you can do:
use B::TerseSize;
use Devel::Symdump;
my $stab = Devel::Symdump->rnew("main");
my %
On 9 Apr 2009, at 15:08, Edmund von der Burg wrote:
2009/4/9 Andy Armstrong :
That sounds as if it could just be Perl getting up to cruising
(memory)
altitude. Perl isn't especially keen on giving memory back to the
OS once
it's used it; instead it keeps it hanging around and uses it to
s
On Thu, Apr 09, 2009 at 10:27:57AM +0100, Edmund von der Burg said:
> What is the equivalent for memory usage?
Devel::Gladiator?
On Thu, Apr 9, 2009 at 2:08 PM, Edmund von der Burg
wrote:
> 2009/4/9 Andy Armstrong :
>> That sounds as if it could just be Perl getting up to cruising (memory)
>> altitude. Perl isn't especially keen on giving memory back to the OS once
>> it's used it; instead it keeps it hanging around and use
On Thu, Apr 09, 2009 at 02:08:26PM +0100, Edmund von der Burg wrote:
> PID USER PR NI VIRT RES SHR S %CPU %MEMTIME+ COMMAND
> 2050 web 15 0 126m 112m 5624 S0 1.4 0:21.66
> /usr/sbin/apache-perl
> 1702 web 15 0 126m 112m 5420 R6 1.4 0:22.88
> /usr/s
Andy Armstrong writes:
> On 9 Apr 2009, at 14:08, Edmund von der Burg wrote:
>> Do these numbers look reasonable to others? The SHR value seems crazy
>> low, but that might be me reading it all wrong.
>
>
> SHR is AFAIK memory that's been explicitly shared and excludes memory
> that just happens
On Thu, Apr 9, 2009 at 2:08 PM, Edmund von der Burg
wrote:
> Perhaps you're right - but our cruising altitude would seem to be very
> high: our biggest processes according to top:
I'll take a guess that you're using Redhat, or some derivative
thereof. The perl included with that includes a lot o
Edmund von der Burg wrote:
2009/4/9 Andy Armstrong :
That sounds as if it could just be Perl getting up to cruising (memory)
altitude. Perl isn't especially keen on giving memory back to the OS once
it's used it; instead it keeps it hanging around and uses it to satisfy
future allocations.
Quoting Edmund von der Burg :
I don't think that it is a memory leak - the size tends to remain
constant after a few requests (it's a webapp - Catalyst under
mod_perl).
Run 'ab' against your app to check that is indeed the case.
More info at http://modperlbook.org/html/14-2-6-Memory-Leakage.ht
On 9 Apr 2009, at 14:08, Edmund von der Burg wrote:
Do these numbers look reasonable to others? The SHR value seems crazy
low, but that might be me reading it all wrong.
SHR is AFAIK memory that's been explicitly shared and excludes memory
that just happens to be shared by COW.
Those figur
2009/4/9 Andy Armstrong :
> That sounds as if it could just be Perl getting up to cruising (memory)
> altitude. Perl isn't especially keen on giving memory back to the OS once
> it's used it; instead it keeps it hanging around and uses it to satisfy
> future allocations.
Perhaps you're right - but
On 9 Apr 2009, at 12:31, Edmund von der Burg wrote:
2009/4/9 Joel Bernstein :
Can you give some more information as to the nature of the code and
its
naughtiness?
I don't think that it is a memory leak - the size tends to remain
constant after a few requests (it's a webapp - Catalyst under
On Thu, Apr 09, 2009 at 12:31:08PM +0100, Edmund von der Burg wrote:
[...]
> I don't think that it is a memory leak - the size tends to remain constant
> after a few requests (it's a webapp - Catalyst under mod_perl).
I'm not convinced that mod_perl is a good environment to run web apps as it
make
On Thu, Apr 9, 2009 at 12:40 PM, Andy Armstrong wrote:
> That sounds as if it could just be Perl getting up to cruising (memory)
> altitude. Perl isn't especially keen on giving memory back to the OS once
> it's used it; instead it keeps it hanging around and uses it to satisfy
> future allocation
On 9 Apr 2009, at 12:31, Edmund von der Burg wrote:
I don't think that it is a memory leak - the size tends to remain
constant after a few requests (it's a webapp - Catalyst under
mod_perl).
I reckon that most likely we're doing lots of innocent little bits of
in-memory caching which all add up
2009/4/9 Joel Bernstein :
> Can you give some more information as to the nature of the code and its
> naughtiness?
I don't think that it is a memory leak - the size tends to remain
constant after a few requests (it's a webapp - Catalyst under
mod_perl).
I reckon that most likely we're doing lots
On 9 Apr 2009, at 10:27, Edmund von der Burg wrote:
I have some Perl code which is being naughty with its memory.
If this were a speed issue I'd throw Devel::NYTProf at it.
What is the equivalent for memory usage?
Do you suspect that it's leaking or just using lots of memory?
--
Andy Armstr
On 9 Apr 2009, at 10:27, Edmund von der Burg wrote:
Hello,
I have some Perl code which is being naughty with its memory.
In a leaky sense?
If this were a speed issue I'd throw Devel::NYTProf at it.
What is the equivalent for memory usage?
Devel::LeakTrace, Devel::Cycle, etc for finding m
Hello,
I have some Perl code which is being naughty with its memory.
If this were a speed issue I'd throw Devel::NYTProf at it.
What is the equivalent for memory usage?
Hugs Kisses,
Edmund
--
e...@ecclestoad.co.uk - http://ecclestoad.co.uk
Help Perl newbies: http://www.send-a-newbie.com/
20 matches
Mail list logo
