Answer the question
In order to leave comments, you need to log in
How to manage CPU resources if all servers on the host are allocated the maximum number of CPUs?
Situation - there is a virtual server on windows 2012 on VMWare ESXi, with 8 processors and 98% of the time this is enough, but from time to time the server is heavily loaded and requires more processors. Solved by adding processors and rebooting. There are other servers on the host.
1. I wanted to somehow solve the problem on a hot one and add resources to the server without rebooting. Of course, you can fix the code, find the cause of the high load, but at the moment this is not possible. The question is about vmware esxi.
2. Let's say that all servers on the host are allocated a maximum of processors. 10 servers, all have 10 percent. How will the load be distributed and how can this be managed in such a configuration? Servers do not need 10, but sometimes you need to add it when the load is high and you need to do this without detriment to other servers.
Answer the question
In order to leave comments, you need to log in
1. In the virtual machine settings there are checkboxes "Allow CPU hot add" and "Allow memory hot add". Turn off the virtual machine, check the boxes, turn on the virtual machine. After that it will be possible to add processors. It is impossible to remove it, it is possible to reduce the amount of allocated resources only with a system reboot.
2. CPU resource planning in VMware is not an easy topic. If you are interested in giblets to look, then here: https://www.vmware.com/content/dam/digitalmarketin...
In general, if the processors are not loaded, then there are no particular problems, because Varya will plan the distribution of the CPU as it loads normally. But as the load increases, virtual machines can start to get dumb, despite the large number of dedicated vCPUs. If there are many VMs on the host with a large number of vCPUs, then virtual machines can get up in the Co-Stop state - a pause in planning for the execution of some vCPUs due to the inability to ensure their synchronous operation within a separate VM, because the number of free physical cores is not enough for simultaneous cycle, and the virtual machine is required exactly at the same time for this particular operation.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question