Win32 heap and <Other>

Use this forum to discuss subjects that don't belong in the other forums.
Post Reply
cyberjawsB

Win32 heap and <Other>

Post by cyberjawsB » Tue Jul 26, 2005 8:29 pm

I have noticed that in my app the win32 heap and the <other> category seem to be the ones that are causing the leak. What exactly composes the information attributing to the <other> category, and do you have any suggestions as to how to begin to decompose these categories to find a specific culpret?

Cheers,
-cj

Andreas Suurkuusk
Posts: 1000
Joined: Wed Mar 02, 2005 7:53 pm

Post by Andreas Suurkuusk » Thu Jul 28, 2005 6:34 am

The "Other data" and "Win32 heaps" nodes contain memory that we have not been able to identify (i.e. it's not part of the managed heaps, or a module). It’s possible that the unidentified memory is used by unmanaged resources wrapped in managed classes (which are usually disposable). If you use such classes (bitmaps, database connections, etc.), I recommend that you make sure that they are correctly disposed (using the dispose tracker).

Unfortunately, the current version of the profiler does not present unmanaged resources used by the profiled process. We’re currently working on an unmanaged resource tracker, which will be included in v3.0 of the profiler. This resource tracker will give you more detailed information about the native memory usage of your application, but currently we only provide detailed information about the managed heaps.
Best regards,

Andreas Suurkuusk
SciTech Software AB

arif
Posts: 1
Joined: Wed Oct 10, 2007 8:44 am

Post by arif » Wed Oct 10, 2007 8:50 am

Hi,

I'm having same kind of problem. profiler reports that <other> category is the one causing the problem. now under this category i see another <other> where all the memory is used up.

how can i know which managed resource is using all this <other data>. e.g which managed data type is responsible for this unidentified memory. my application doesn't uses any unmanaged resource directly.

thanks
arif

Andreas Suurkuusk
Posts: 1000
Joined: Wed Mar 02, 2005 7:53 pm

Post by Andreas Suurkuusk » Wed Oct 10, 2007 8:53 pm

No managed instances are directly responsible for the memory usage presented as <Other data> (since their memory is presented under the Managed heaps nodes). However, they might wrap unmanaged resources that are not correctly released. Using the dispose tracker and the unmanaged resources tracker might provide you with useful information about this. You can read more about the <Other data> node at:

http://www.scitech.se/blog/index.php/20 ... ther-node/
Best regards,

Andreas Suurkuusk
SciTech Software AB

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest