Hi AMD Server guru's ,
I need your assistance for the below, we purchased a Thread ripper in June 2020 from newegg.ca. Manufacturer: Gigabyte Technology Co., Ltd. Product Name: TRX40 AORUS MASTER
1. How to utilize all the cores in this Model: AMD Ryzen Threadripper 3990X 64-Core Processor (Ubuntu 18.04 LTS installed , on top of that using LXC containers to build the virtual machines) 128 CPU , 256GB RAM and 1.8TB nvme disk . Seems like 8 CPUs it is using while doing the builds inside the server.
2. How to remove the parallel restriction and cleaning the shared state ?
We are trying to build a vm inside the threadripper to monitor the build time speed.
While doing build it seems Ubuntu server OS is not utilizing the whole cpu, just using 8 cpus only. Is there any chance to use the whole cpu power at a time and reduce the build time.
Thanks in advance
You have to wait for the kernel of your operating system to update for your processor.
srinig, please see my reply to this thread. Good luck and enjoy, John.
I'm an Unix expert. My specialty is server system installations. Apart from LXC, I know 2 more types of virtualization and I can configure any server. But I have no money to buy a Threadripper.
Edit: LXC kernel is installed independently of the operating system. OP may need to wait for the LXC to update.
Edit: LXC and OpenVZ contain a very similar virtualization technology. But I'm not sure which one gets more frequent updates. OpenVZ can also be tried if you are not using a custom control (or something as) for LXC.
I recommend posting your query in the correct forum here: https://community.amd.com/community/server-gurus
Sorry for the trouble , I find the LXC has 4 containers on that build server. these containers has priority to steal resources from developers Depending on the build, not all cores would be used because of recipe ordering and dependencies.
LXC virtualization provides the necessary support for CPUs with high core counts. These changes are made with the command system or a control panel. You can solve your problem by doing a little Google search. If your OS doesn't see the cores of your processor, a kernel update is required. If your OS sees the cores of your processor but doesn't use it, you need to check the settings in the kernel or the containers.