Hi guys.

 

I'll jump right into it. I have 16 GB of ram. I run a windows VM that's always on that get's 8 of them. Generally I don't have issues with the other 8 filling up, but lately they have, which crashes my VM when I run out.

 

The thing is, I don't really run much else. I run crashplan and syncthing in docker containers, and usually plex and transmission but I've turned the latter two off since this problem started.

 

My VM always crashes when crashplan is running a backup.

 

Even after the VM has been killed, my ram usage hovers above 50%.

 

If I stop all of my docker containers, including crashplan, my ram usage still stays where it was after the VM was killed.

 

I have observed this behavior when I know for a fact nothing should be accessing my shares or doing anything else. With no containers and no VMs running, and none of the clients on my network accessing the shares, unraid should just sit there and do nothing, but memory usage stays where it was when it crashed my VM.

 

How can I go about diagnosing this? I get a feeling it's a prerelease bug but I don't know how to find more information to determine if it is or isn't more conclusively.

 

If it's not I'll repost this to a different sectiion.

 

(i'm getting my ram usage information from the unraid webui dashboard tab, lmk what else I can do to get more detailed memory usage info)

I would recommend logging in using ssh and running  the "top"  command to show what everything is using then post a screenshot here.  There is a way to order the list by memory usage but I can't remember how off the top of my head (a quick Google will say how)

 

Jamie

Crashplan is (to put it mildly) a pig on memory usage.  The more it has to backup, the more memory it uses.  Don't know if it's properly releasing the memory however on its stopping.

 

You can limit the amount of memory that any container is able to use.  Look in the docker FAQ for this.

 

Also with 8 gig memory allocated to the VM, after overhead taken up by KVM, etc you're probably looking at ~10gig used for the VM

 

How can I go about diagnosing this? I get a feeling it's a prerelease bug but I don't know how to find more information to determine if it is or isn't more conclusively.

Unless you have evidence that this doesn't happen when you run the latest stable I don't know how you can support the idea that it is a prerelease bug. If you don't have this evidence I will move the thread to General Support.

 

*edit*

If it's not I'll repost this to a different sectiion.

Please don't repost it. I can move it.

How can I go about diagnosing this? I get a feeling it's a prerelease bug but I don't know how to find more information to determine if it is or isn't more conclusively.

Unless you have evidence that this doesn't happen when you run the latest stable I don't know how you can support the idea that it is a prerelease bug. If you don't have this evidence I will move the thread to General Support.

 

*edit*

If it's not I'll repost this to a different sectiion.

Please don't repost it. I can move it.

 

Thank you.

 

The only evidence I can offer is that this combination of docker containers and VM has lived happily together for months on the stable release. That being said, I've also had the RC since December and this just started happening this week, so I can't conclusively point to it being related. Also from reading the other replies it sounds like it may just be a crashplan or docker problem, not an unraid problem. Perhaps a move would be best

How can I go about diagnosing this? I get a feeling it's a prerelease bug but I don't know how to find more information to determine if it is or isn't more conclusively.

Unless you have evidence that this doesn't happen when you run the latest stable I don't know how you can support the idea that it is a prerelease bug. If you don't have this evidence I will move the thread to General Support.

 

*edit*

If it's not I'll repost this to a different sectiion.

Please don't repost it. I can move it.

 

Thank you.

 

The only evidence I can offer is that this combination of docker containers and VM has lived happily together for months on the stable release. That being said, I've also had the RC since December and this just started happening this week, so I can't conclusively point to it being related. Also from reading the other replies it sounds like it may just be a crashplan or docker problem, not an unraid problem. Perhaps a move would be best

 

Set the dockers to not autostart, reboot and see if it happens again.. I am running crashplan and a lot more and do not notice anything wrt memory usage..