I have a program that does a lot of file IO, multiple DBI DB access,
even spawns a perl program and uses that output. The program runs in
about 7 minutes. Where I add require “profile” to the main file, the
program can be running for hours and still not finish.
I killed the program and removed the file writing, there is still a
lot of file reading, and I removed the system calls that spawn a perl
program, but it still takes hours and I end up killing the process.
I have looked around but could not find any major “problems” with
running the profiler on the web. Are there certain types of calls
that are known to cause issues with the profiler?
Any insight would be appreciated.
Thanks,
Walt
···
Walter Szewelanczyk
IS Director
M.W. Sewall & CO. email : walter@mwsewall.com
259 Front St. Phone : (207) 442-7994 x 128
Bath, ME 04530 Fax : (207) 443-6284
Welcome to the world of "proof-of-concept" ruby tools.
The profiler and the debugger are implemented in an extremly bad ways.
It is not a problem to increase the performance by a factor 100.
I'm currently working on the debugger issue. Profiler will come later.
At the moment the only possibility is to reduce the data size in your
application or split it into a few testcases. I know it removes most of the
usefullness of profiling, which should be done on real world data. But
at the moment there is no way to do this.
Welcome to the world of “proof-of-concept” ruby tools.
The profiler and the debugger are implemented in an extremly bad ways.
It is not a problem to increase the performance by a factor 100.
I’m currently working on the debugger issue. Profiler will come later.
At the moment the only possibility is to reduce the data size in your
application or split it into a few testcases. I know it removes most of the
usefullness of profiling, which should be done on real world data. But
at the moment there is no way to do this.
Isn’t your company trying to build a ruby debugger to sell?
Isn’t your company trying to build a ruby debugger to sell?
If he is, is that a bad thing?
No, not in theory. However one might be a bit biased about a free
debugger if one is trying to sell another debugger. I will have to
agree that the debugger needs help though (however I don’t want to have
to pay for it so I’ll hope for improvements to made to the current one).
An example is below; every now and then I get a mail that shows the
last “n” bytes of a message repeated. (n ~ 45)
Is this a known issue of … the mail/news gateway, or something
else? I see it in both a web interface to yahoo, and when I POP the
mail to a local mail client.
An example is below; every now and then I get a mail that
shows the last “n” bytes of a message repeated. (n ~ 45)
Is this a known issue of … the mail/news gateway, or
something else? I see it in both a web interface to yahoo,
and when I POP the mail to a local mail client.
My guess is that the issue is on your side, as this is the message I
received:
mgarriss wrote:
Isn’t your company trying to build a ruby debugger to sell?
Is this a known issue of … the mail/news gateway, or
something else? I see it in both a web interface to yahoo,
and when I POP the mail to a local mail client.
My guess is that the issue is on your side, as this is the message I
received:
mgarriss wrote:
Isn’t your company trying to build a ruby debugger to sell?
If he is, is that a bad thing?
Could it be something on your mail server?
Entirely possible… It just started in the last couple days, and it happens
only on 1 out of every 10-20 mails. Odd.
Guess I’ll live with it; was just wondering if anyone else saw it.