I’ve noticed lately ashServ.exe memory usage in Task Manager gets rather high at times (I’ve seen it get as high as 54,000+kb!!!) and does not appear to release it for the remainder of that pc session. So I did some curiosity observing today.
Booted machine, signed onto WinXP (on my LUA, as always) and opened Task Manager. First screenshot shows typical Avast memory usage at boot up. no CPU usage showing for ashServe.exe. Left TM on my screen.
Scheduled 9AM SuperAntispyware full scan began. I noticed ashServ.exe memory usage begin to climb, and CPU usage raged from 4%-13%. FYI I only run Network, Standard and Web shields.
Mid-SAS scan I stopped my 3 on-access Avast scanners and the memory usage stopped climbing at around 30,000kb. Stayed steady until I turned my 3 shields back on, as I would expect. Then resumed its relentless increasing.
At the end of my SAS scan, ashServ.exe had climbed to the number in my second screenshot and stopped increasing, as I would expect.
But I waited quite awhile and that memory usage number did not then drop back down to the lower usage figure, which I WOULD expect. If I were to then go on-line with my default browser (FX 3.0.8) that has known increasing memory usage, coupled with what ashServ.exe doesn’t release, well…you get the picture. Surfing slows down quite a bit the longer I’m on-line.
Can someone tell me why ashServ.exe doesn’t release the memory it is no longer using? Is this an Avast problem or a WinXP problem? As I only have 512MB RAM, this can be an issue at times.
Well, I stand corrected & have to eat my words. :-[ It didn’t release the memory right away, but after going on-line for a little while, I see it did finally release that memory, per the atttached screenie. Consider this inquiry resolved.
Memory is to be used… this was discussed some times in the past. The value to be seen, if any, it the Virtual Size (a hidden column into the Task Manager) and not the Mem Usage column. When you’re using SAS, it’s normal that the resident avast scans the files being accessed. Some users say to disable the antivirus (Standard Shield) while running other on-demand scanning.
Whoa! You’re right, Tech! I just revealed my virtual memory column and look at how much SAS can use at will (I assume during updates and scans). I’ll have to repeat my usage observations but this time watch the Virtual Mem column. That’ll give something to do this afternoon :). (We retired folks have to look for inane things to fill our time! ;D ). Actually I worked my tail off planting a bunch of new grass in my back yard, so I’ve earned the right to play on the computer the rest of the day.
Interesting read. Guess I looked over that thread in my search results. Their testing was far more extensive & technical for me. Was just curious as to whether my situation was unique and indicative of some action I needed to take to correct. It would appear what I’m seeing is normal for Avast. Whatever the technical causes, I like Avast too much to ever switch to another camp. So I can just live with how ashWeb.exe and ashServ.exe manipulate physical/virtual memory. It will be interesting to see how the new Avast v.5 handles memory.
XP Task Manager ™ and Process Explorer (PE) use different terminology/meanings for memory usage (Vista TM basically being the same as PE). By understanding the terminology and measurements in TM we can see if leaks are present and to what extent.
In XP TM, private bytes are the current number of bytes a process has allocated that cannot be shared with other processes—in other words, how much nonshared memory your process is using. This memory is more important than shared memory when multiple processes execute on a machine and are the primary counter for measuring memory leaks and overall memory usage. This corresponds to the VM Size column in TM. Some of this memory could be swapped to disk at some point and not affectting other apps in RAM. Still, this is the amount of memory the app uses in a non-shared way and is important to minimize and if it grows appreciably over time it represents a memory leak.
Mem Usage in TM refers to the number of pages of memory committed to a given process, both shared and private - also called the working set. This is what is being kept in RAM at the time it is measured. This will fluctuate over time as compared to VM Size which shouldn’t. Since shared memory is included, if another app does need the memory space it will be released so remaining size is not as big an issue.
You can use perfmon to measure private bytes and working set and graph out to see application memory performance.
Thank you for taking the time to reply, Reductase. Your explanation about Virtual Memory was clear, succinct and most helpful. I just booted up with the VM column showing in TM and watched as I went from website to website. Avast’s VM numbers sat perfectly still: ashServe.exe around 29,700K; ashWevSv.exe around 27,100K. Apparently no leaks there.
Firefox’s MemUsage and VM, on the other hand, both climbed steadily with every website change. But we all know about FX well-documented memory leaks and Mozilla forum discussions on said topic, don’t we? Sure wish the FX developers could get a handle on that issue. Most likely THAT is at the root of my slow down when I stay surf the net a very long time, not Avast.
Again, thanks for the mini lesson in non-technical terms. This retired teacher greatly appreciates simplicity.
Was just curious as to whether my situation was unique and indicative of some action I needed to take to correct. It would appear what I’m seeing is normal for Avast. Whatever the technical causes, I like Avast too much to ever switch to another camp. 