CentOS OpenVZ container clone RAM is reduced to minimum after rebooting.

1 post / 0 new
#1 Tue, 11/03/2015 - 01:19
devteam

CentOS OpenVZ container clone RAM is reduced to minimum after rebooting.

I have Cloudmin licensed and installed. I also have Virtualmin/Webmin licensed and installed as one Virtual server. This is an OpenVZ container running just fine and I simply want to make a clone of it.

I execute the steps to create a clone, make sure I am going to assign it like all of the other container installs with 4GB of RAM and 4GB of swap. The swap is likely not needed with that sort of RAM available, but humor me. After a couple days worth of noodling, I figured out "when" the RAM is taken away from the container and replaced with 256MB of RAM instead. It is fine so long as I do not reboot the cloned container or the entire server. But upon any sort of reboot, that container (and not the identical container it was cloned from) changes the RAM allotment from 4GB to 256MB. I am familiar with OpenVZ and at one point or another ran all of the commands that should have placed the system with 4GB of RAM, rebooted the container and it failed.

So what am I missing? I am really doubting it to be a setting I have set myself, but in resources if I change the "Guaranteed memory allocation" off of its default "Same as maximum memory" and insert any number in there, it changes nothing and the entry I made for "Guaranteed memory allocation" has gone back to its default.

I am really convinced after two days of this crap it is a genuine bug of sorts. Now whether or not it is a mishap on this installation, just an unlucky collision of all the things it takes to run a virtualized system, I do not know. I can't point my finger at the software alone as I guess the only way I would know for sure who and what is responsible would mean an entire fresh install just to see if it still has this bug to it. I don't find that appealing at the moment... Can anyone shed some light on this topic? I have combed the settings several million times and it creates the clone every time properly, but a reboot always drops all the RAM. Could it be messing with me because I am cloning licensed Virtualmin container? I figure one of two things would happen, A) I would be asked for a license, or B) be told I need to downgrade that clone, which is fine with me, but it needs RAM, license or not. It is only going to mainly be a glorified DNS server anyhow, but will not at all run on 256MB of RAM. Thanks in advance.

In hindsight, it would not be too shocking to think it is an error caused due to another problem that I gave up on. Cloudmin has a memory allocation error with any and all OS images except for those that it "ships with" available from you folks. This means since I am running CentOS 6.7, and your images are of 32bit 6.0, it is a long and retarded journey to update the OS to the current CentOS 32-bit version. This issue - where the images refuse to install in the first seconds of execution based on an error that indicates there is not enough RAM and "unexpected Integer" (as in the error message it indicates the RAM available is 0.0xxxxxxx for whatever reason. So this is a second issue, but I guess I am saying since I had to spend hours updating and correcting stuff to move from CentOS 6.0 you offer and 6.7 where it wants to be, is a horrific and tedious situation too and very well may be how I arrived here, but I don't know why the container I am cloning is working fine with no issue??? I am baffled. I will throw up another post here about the inability of Cloudmin to use and images but those available within to download.

Douglas