I have noticed that one of my containers uses much more SWAP memory than allocated. Allocated are 512MB and it uses well over 700MB, although there is still plenty of RAM free. Can anyone give me a hint why?
Now the SWAP of this container is back to 0% and the container uses 34% of its RAM.
On this container runs a vesta with nginx. If I now call a website that runs there, the SWAP goes directly and slowly up again. In addition, if I start another VM on the node, the SWAP directly rises to over 100% from the already mentioned container.
Now the SWAP of this container is back to 0% and the container uses 34% of its RAM.
On this container runs a vesta with nginx. If I now call a website that runs there, the SWAP goes directly and slowly up again.
This looks like normal behavior to me. LXC Containers use the swap from the host, and the only way to completely avoid swapping is to disable swapping on the host.
The swap rising over 100% is because of how swapping in cgroups implemented. "memory" is actually RAM + swap in cgroups version 1. When you add a swap to your container to use, it adds extra swap space on top of the current memory + swap.
No its not... There is a windows 10 running and the vm is normally turned off. I booted it to test it. After setting vm.swappiness to 1 this seems not to be the case anymore.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.