Your data. Anywhere you go.

New Relic for iOS or Android

Download on the App Store    Android App on Google play

New Relic Insights App for iOS

Download on the App Store

Learn more

Close icon

Server Monitor High RAM Usage - POSSIBLE BUG?



The server monitor, after running for 2 weeks or so will start using in excess of 2GB of ram and will stop reporting to new relic. I have to manually restart the server monitor to get it to report again. After a restart, the ram usage goes to around 60MB.

I could solve this issue by creating a scheduled task to restart server monitor daily, but it seems like something that shouldn’t have to be done.

I’m using the latest server monitor for Windows Server 2008 R2 - which is version I’ve been seeing this issue across more than a few of my servers so I’m guessing this can’t just be a coincidence. Has anyone else had this issue on Windows Server?


Hi @m18488482,

I agree, scheduling a task to restart the monitor is not ideal. Let’s figure out what is causing it.

  • Do you see any increased CPU usage?
  • A warning in the Application Events Logs containing a description of Process sampling is taking over 15 seconds. ?


I am also seeing this issue across many servers in my fleet running WIndows Server 2008 R2. New Relic will stop reporting before Memory becomes excessive, so it is difficult to see the history of CPU usage. I didn’t see anything suspicious in the Application Events Logs. On some servers, New Relic Server Monitor was taking more than 3GB of memory


I am relieved to hear I’m not the only one experiencing this… A few colleagues of mine have had this happen as well.

I don’t see any increased CPU usage or any errors in the event viewer about process sampling. This is continuing to happen months later as well.


@m18488482, @robertg - This is a known but relatively rare issue. We thought it might be related to the another issue (hence the question about the “process sampling”). We have a bug filed for it and I’ll add your voices to the bug report. Unfortunately, we don’t have a lot to go on at this point and we haven’t been able to reproduce it ourselves. Restarting the service periodically is your best bet at this time.

@m18488482, @robertg mentioned seeing this on his fleet running WIndows Server 2008 R2. What system(s) do you see the issue on?


@Tjd This might provide some insight into this issue but not sure. I’ve switched to VMware ESXI from Hyper-V and this issue isn’t happening anymore. It also appeared to happen more frequently on Hyper-V with dynamic memory turned on. I’m not sure if this has anything to do with this problem or if its just a coincidence since it happens randomly.

I was seeing this issue only on my Virtual Machines that were Server 2008 R2 servers (I haven’t seen it yet on my 2012 servers).


Thanks for sharing your discovery, @m18488482. We’ve added this information to the report.


I’m still seeing this issue across my 2008 R2 Servers. In my last job, I ended up reverting to an older version of the client which resolved the problem. I thought this would have been resolved by now ?


@niaccarino - Unfortunately, this has not been resolved yet. I wish I had better news. It sounds like downgrading is the solution that works best for you. A periodic restart of the service also works but of course is less than ideal.


This problem started happening for us this weekend, too. I’ve opened a case with NewRelic support.

It affected 3 of our 7 application servers and caused the applications on those servers to become unresponsive. Server Monitor was utilizing around 3GB of RAM! I hope New Relic understands the severity of the issue, though it’s not a good sign that this issue was first reported in July 2015 and it still has not been resolved. If this cannot be resolved, we may have to stop using New Relic. I would not consider scheduled restarts of the service an acceptable solution.

We are running on Windows Server 2008 R2 as well. CPU usage is normal, RAM is high, and the server monitor had stopped reporting.


Great thanks! Ill continue using the old agent.


Experiencing the same issue on multiple servers. All of a sudden “New Relic Server Monitor” gobbles up all available ram. Restarting the service solves the problem. We’re running but I cannot see anything in release notes that this is fixed in later versions? Which version have you downgraded to?

If this is not solved we cannot continue to use New Relic



We will need more information from your system to provide you a better answer. I have created a ticket on your behalf.




Hi. Was there a fix for this issue? We just ran into exactly the same problem running Windows 2012 R2 New Relic .NET Agent (64-bit) New Relic Server Monitor was taking up more than 4 gigs of memory. Had to restart the service to fix it.


Hi, @jchen1: Are you sure it is the Server Monitor? That product has been retired and should no longer work.


Yes. I am very sure.


The Windows Server Monitor service may run but New Relic is no longer collecting data from it. I would suggest uninstalling it.


Hi @jchen1 - to add some context to @Tjd’s update, we announced last summer that we would be bringing our Servers product to end of life, and it was decommissioned on May 15 of this year.

As Ted mentioned, the agent is still there if you did not uninstall it, but we are no longer collecting data from the agent. We apologize if our messages did not reach you and this is catching you by surprise. Please let us know if you have trouble uninstalling the agent.


I uninstalled the service. Thanks all for your help


Thanks @jchen1! We’re here if you need anything else.