Tarun Posted November 9, 2006 Posted November 9, 2006 LLXX, there's no need for citation when biznatchio not only wrote the article, he also knows plenty about Windows memory management.That's what you think... according to you someone could write a huge, verbose, official-looking essay claiming anything and you would probably believe it because the author "seemed to know much on the subject".Having no references makes it all the more suspicious.After searching the Internet a few times to see if this information has been published somewhere else, (e.g. http://www.google.com/search?q=%22task+man...art=10&sa=N ) there is nothing else that supports the claims found in your source.From what I've seen, Task Manager is reporting accurately. See http://www.msfn.org/board/index.php?s=&...st&p=581316 for one example of calculating a memory usage. Perfect example is with the Hosts file. Just because a multitude of websites say to use it to block malware and advertisements doesn't mean the information is correct. I'll look into it more to see if I can prove that it is indeed correct; though just reading it is more than enough because it makes perfect and practical sense.
CoffeeFiend Posted November 9, 2006 Posted November 9, 2006 (edited) Actually, it's common knowledge that the task manager isn't exactly a great way to measure memory usage. Some of the columns are actually mislabeled/quite misleading (like private bytes is labeled "VM size", "mem usage" is really the working set, etc). There are resources out there (by experts in the subject matter) that explain all this in great depth. Process explorer is by FAR a better tool (shows all you could ever want to know, and labeled properly!) Or windows' own performance counters even (they're labeled properly, not matching with task explorer's)Your method is perhaps less accurate than the task manager - close to 100% wrong. You're accounting for the exe "image size" (a very small portion of the equation), but then go and totally ignore any kind of memory allocation (making this method laughable at best), and then counting DLLs it loads towards memory used - even though they're NOT loaded twice (if any app is already using them, they're NOT loaded a 2nd time, Windows is smarter than that). And telling yourself your math is right because the number (a number too small to really verify anything) "seems" to match what's really the working set. You sir, after nitpicking at a fairly decent article, have managed to produce the most laughable attempt at calculating memory usage I've EVER seen. By your perfect method, a 10kb program that would malloc a terabyte of RAM would use like a megabyte or so (mostly because of dependancies/DLLs which are already loaded, which really require no memory).Edit: And here, before you call me a liar: a quick 3min25s sample video (by sysinternal experts Mark Russinovich & David Solomon) that will tell you it's true the columns in task manager are misleading. Or an excerpt from a MSDN blog (right out of the horses' mouth, in case you think the sysinternal folks are n00bs, even though they work for MS now):most people use task manager to look at memory usage of a process, and specifically look at the Memory Usage column. Surprise surprise:) what this actually shows you is the working set of the process...If you want to see private bytes which is a far more interesting counter, you sould look at the column in task manager that is labeled Virtual Memory Size (yeah, that's really intuitive:))This is explained on hundreds of pages, in various programming books, newsgroups, etc. And not by random nobodies either.When using the proper columns, the task manager can give an approximation/general idea of how much memory an app uses, but it's not exactly a "accurate to the byte" meter. Edited November 9, 2006 by crahak
Jeremy Posted November 9, 2006 Posted November 9, 2006 (edited) I was talking about v2.51. The later versions are getting worse NOD32 v2.51.30 IdleNOD32 v2.51.30 (After Scanning with its maximum capability)Approx:Before Scan:Mem Usage - 21,556Peak Mem Usage - 33,004VM Size - 18,468After Scan:Mem Usage - 44,788Peak Mem Usage - 68,924VM Size - 37,780Approx:Before Scan:Mem Usage - 23,124 MBPeak Mem Usage - 29,132 MBVM Size - 19,428 MBAfter Scan:Mem Usage - 43,848 MBPeak Mem Usage - 62,768 MBVM Size - 37,644 MBAs you can see, generally speaking, there isn't a significant difference between v2.51.30 and v2.70.90Again, I realize that the TM's numbers change when you simply move a window around or whatever, so obviously they aren't exact, just generally giving you an idea. Edited November 9, 2006 by Jeremy
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now