Anti-Spam Discussion Forums

[spaminator]

I ran spamitback all day today. The program had consumed all memory and processing power. When I could finally get Process Explorer up this is the info I got on spamitback.

Look at the page faults, the virtual size, and peak working set values for the second instance of spamitback. I have a gig of RAM in this thing, and as always this program consumes it all.

Process PID Handles CPU Time Threads
spamitback.exe 920 34 0:00:00.421 2
spamitback.exe 3924 300 9:24:16.296 6

Page Faults Virtual Size Working Set Peak Working Set
2,641K 18,616K 276 K 1,556 K
1,930,432,844K 942,440 K 52,564 K 877,240 K

[spaminator]

so Spamitback was this information useful??

[oli32de]

Hi SpamItBack,
I guess you programmed your program using C/C++ so I searched the web and found an interesting site. This guy tested so-called memory checkers that are able to find "misplaced" pointers that might 'cause our SpamItBack problem regarding memory leakage.

[quote]
http://www.cs.utexas.edu/users/jpmartin/memCheckers.html
[/quote]

BTW, how about a possibility to run SpamItBack with a parameter (e.g. spamitback.exe -log) that starts logging of information so it might be easier to report a bug or show a certain behaviour and you also have a "proof" and maybe an idea, what might have caused the problem ...

CU

[FORT1]

I am a C/C++ programmer myself. The rule is very simple. If you create an object (generally with "new"), you must destroy it as well ("delete").
I have never had any problems with memory leaks.

If I were you, I would simply browse through the code, examining all the "new" statements and checking if the objects are properly destroyed.

ClassA* obj = new ClassA();
...
...
delete obj;
obj = NULL;

Ps. Dear Mr. Spamitback, please hurry! "My" spammers need some attention urgently.