VM Sweet Spot Config
Posted: Sat Apr 04, 2020 6:40 pm
My home lab has a server I use to host various VM's. At the moment I have some unused capacity to dedicate to FAH. The server is a SuperMicro rack server with dual Xeon X5650 (12 cores) and 72GB of RAM. I know this isn't a very powerful box but every little bit helps.
I started with one VM with one core and 1 GB of RAM. Once I got it running I cloned it and started another VM with 2 cores and 8GB of RAM. Both systems only used a peak of 50% CPU until I added a second CPU slot to each. Now they both run at 100% CPU. Neither system is using much of the RAM provided. The single-core 1GB VM reports just over 2000 PPD while the dual-core 8GB VM is doing around 7000 PPD.
I'm trying to find information about how to configure a given VM for the best return. Do I add cores to these VM's or spin up more VM's. Memory doesn't seem to be a factor but there must be a minimum.
I'm hoping to find the sweet spot for a few small systems instead of one big one. Then I can scale capacity as needed. I have a Docker Swarm. Has anyone tried working with containers?
Thanks,
AustinPowered
Team: Linux For Everyone
I started with one VM with one core and 1 GB of RAM. Once I got it running I cloned it and started another VM with 2 cores and 8GB of RAM. Both systems only used a peak of 50% CPU until I added a second CPU slot to each. Now they both run at 100% CPU. Neither system is using much of the RAM provided. The single-core 1GB VM reports just over 2000 PPD while the dual-core 8GB VM is doing around 7000 PPD.
I'm trying to find information about how to configure a given VM for the best return. Do I add cores to these VM's or spin up more VM's. Memory doesn't seem to be a factor but there must be a minimum.
I'm hoping to find the sweet spot for a few small systems instead of one big one. Then I can scale capacity as needed. I have a Docker Swarm. Has anyone tried working with containers?
Thanks,
AustinPowered
Team: Linux For Everyone