Richard Lynch wrote:
On Mon, September 18, 2006 11:54 am, Chris wrote:
I'm hoping somebody can give me advice on how I can go about debugging
a
problem I've been having. The bottom of this email has some of the
details of my system.
There is an intermittent slowness throughout my site. At seemingly
random points in random scripts at random times it just seems to wait.
A few of these have been as long as 600 seconds, just sitting there,
apparently doing nothing. I've tried tracking down the problem using
microtime(), that jsut showed me how (mostly) random it was. The least
complex area in which I was able to isolate a problem was a 15 line
chunk of HTML, with some <?=MY_CONSTANT?> in them.
Are you running any of the cache programs such as APC/Turck/Zend Cache?
I was running Zend Platform, and thus, I believe, Zend Cache, I've
disabled it for now and I don't think it's the cause of my problem..
Or the Zend Optimizer?
So is the optimizer, though I suppose I could try enabling that and see
what happens.
Try disabling those and re-running the tests, just to be certain it's
not the cache acting up.
I'd also suggest trying some other tools to monitor the activities --
just in principle, to be certain that the monitors themselves are
accurate. I have no particular package in mind -- Just that two heads
are better than one when trying to view what's happening at that level
of detail.
That's part of my problem, I'm nto sure what sort of tools I need to be
monitoring what... I just don't know many places my problem *could* be.
My site has a decent amount of includes, it's probably doing 30-40
includes per request, with an occasional require or two.
Just in general, I'd suggest that 30-40 includes per request is a bit
excessive...
Can none of them be combined?
I do intend to streamline the whole site, but it's not feasible to do it
now. I see the includes as being more of a possible trigger than the
full cause of my problem.
I *believe* the big "hit" for an include is in the disk seek, not
reading and parsing some bits of PHP code.
I'm not really great at reading iostat, but to the best of my knowledge
the hard drives are hardly even being used
This is not likely to be at all related to your current issue -- just
a general observation.
Thanks for the reply.
Chris
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php