Answer the question
In order to leave comments, you need to log in
What could be the problem with this error (Out of memory)?
In general, the situation is this. They dragged the Wirth machine TEMPORARY to another iron, deployed it in the Wirth box.
One machine - nextcloud, data for 400GB, accesses based on reading, traffic is about 10mb / s at a constant rate. The image is divided into 2GB, RAM 8GB, stands on a regular disk, 7200 rpm. Busy RAM - 500-700MB. Inside the virtual machine, everything is quietly copied at the speed of a hard drive. That is, from herself - she copies 45-60mb / s to herself.
The second machine - the Azuracast software is deployed, reading from it (configs are copied, cheap) occurs regularly, but the volumes are very small, it costs on a regular SSD. RAM - 2gb, occupied 500mb. The speed of copying inside the virtual machine is about 100mb/s. Only on the same disk deployed 4 more of the same machines. but they don't get this error.
The main question - I attach screenshots of both machines in order.
The second question is why does the virtual machine that is on the SSD (this is the second one in the question) have such a low speed of communication with the disk? From the host system, the disk produces results almost 3 times higher. (of course, the check was with only 1 included this virtual machine)
Answer the question
In order to leave comments, you need to log in
What is the main question? Apache (perhaps not by himself, but by PHP) is trying to eat up more memory from you than the OS can (or wants) to offer him, so oomkiller starts up and kills the most suitable one in his opinion.
I suspect that on the host the memory allocated to the virtual machines is not much less than the total amount, so the yok occurs periodically. In more or less functional hypervisors, you can turn off memory balunning - then the entire volume will be immediately issued to the virtual machine.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question