Ridiculous Modified Memory usage

Hi,

I've noticed lately that the Modified Memory is getting extremely high over a period of time. I can't really tell when this happens since my pc is running for days if a restart is not needed so it's hard to tell what's causing it.

I have this issue for a while now and I made a little research in the past which left me confused on whether the modified memory is reusable or not. Many people said that it is reusable and Windows will utilize it when the RAM is almost full. The problem here is that I'm running 32GB of RAM so it's a little hard to get to a point where windows are starving of RAM... until today where I received the Out of Memory pop-up window, telling me to shut down programs!

I reconfigured the paging file in the past, adding 2000MB (initial) to 8000MB (maximum) paging file for C:\ but it looks like it didn't solve the problem. Maybe that's not enough? I don't want to allocate a huge amount of my C:\ to paging file because it's an 128GB SSD and every GB is valuable. For the same reason I had to disable the Hybernation mode in windows because it was reserving 32GB of the SSD in order to fit all my RAM.

The rest of my HDDs do not have a paging file which I'm not sure if it's good or not.

Does anyone have any idea how to narrow this down?

Thank you in advance!

There isn't enough information available to diagnose the problem. A screenshot as Callender has suggested would help.

Memory usage of chrome is high but not nearly high enough to explain the situation. Such problems are often caused by driver issues but as the paged and non-paged memory is low that seems unlikely.

"Modified" refers to process private use memory that has been trimmed from the working set of the process. This is complex and I will not explain it further. Normally this would be written to the pagefile and converted to standby memory but with such a small pagefile that cannot happen. A larger pagefile would mask the symptoms but wouldn't solve the real problem.

But that isn't the real problem, merely a symptom. The real problem is why this and the commit charge has become so high.

Thank you guys for your replies!

I had to reboot my pc due to updates installation and tweaking settings so the modified memory is gone for now. Sadly, that's before I see the post from Callender so no RamMap feedback for now.

I will keep the Resource Monitor and the RamMap open in my secondary monitor so i can constantly keep an eye on what's going on. As I said in my first post, it seems like it takes several days until I notice this issue. I can't tell if it's something that's build up over time (like 100MB modified mem per hour) or it becomes huge instantly by a specific program or action.

I am not sure what screenshot you mean. If you're talking about a RamMap screenshot, I will provide one as soon as I get a big modified mem again. If you mean something else, please let me know and I'll provide everything!

This observation is intriguing! I was using Event Viewer too when the modified memory issue occurred! i left Event Viewer open before I leave for work and I will check again once I get back to see if the modified memory increased!

Modified memory is pagefile backed memory that has been trimmed from the working set of a process or other entity. There are only 2 way that modified memory can be reduced.

1. The entity from which the memory was trimmed terminates or it frees the address space from which it came. This is complex and difficult to describe in simple terms. Clearing the Event Log might do this in isolated cases. In this case that seems highly unlikely.

2. The modified memory is written to the pagefile and then moved to the standby list. The memory manager normally does this but in this case the pagefile is too small for that to be possible. A larger pagefile would eliminate the symptom (at least in the short term) but it wouldn't solve the real problem.

The question is: from what entity was the memory trimmed? Nothing in the Task Manager screenshot accounts for this. RAMMap should. I can't promise I will be able to interpret the results.

Ok, my Modified Memory has reached ~10GB so I attached a few screenshot for you guys to see the current stats.

I was expecting RamMap to indicate a specific program that's consuming all that memory but instead, it shows almost all of Modified Memory as "Shareable".

I am pretty sure Event Viewer has something to do with this as the Modified Memory tripled since i opened it, but I can't find a direct relation.

I certainly can delete my log files to see if the Modified Memory will decrease but before I do that, is there anything you guys want me to check? As you can see, it took several days to reach this point and if i delete the log files (and actually do the trick), we will need another week or so to get a big Modified Memory again. So if you know any other monitoring tricks (like RamMap), let me know so I can check now!

I'm sorry, i realize this thread is over a year old, but i think i have found a good solution. I stumbled upon this thread because i had exactly the same problem as Evrip. I did not find a working solution here so i continued searching. Then i came across another thread here: https://social.technet.microsoft.com...um=w7itproperf

Someone in that thread suggested the following:

I believe I isolated it to this by adding the Handles column to Task Manager and watching it leak them like crazy, which eventually leads to paged (or non-paged, I forget which) memory spiking and ultimately "low memory" warnings.

I followed the instruction and there was one process that had 90k handles (the next highest was 5k). In my case it was some yahoo toolbar, but other people in that thread report all kinds of processes that caused the same problem for them, but what they have in common is they all found and solved the problem thanks to this one simple suggestion. Once i had found the guilty process, i stopped the service and the total memory usage dropped from 9GB to 5GB instantly. Even after that, the usage still continued to fall slowly, giving me another 0.5GB. I'm so happy i finally found the solution after having this problem for more than a year.

By the way, one other suggestion in said thread was to not only add the handles column, but also the GDI objects column. It might point to the problem as well, but in my case i saw the problem as soon as i looked at handles.