Become a Part of the VMware {code} Community

 

VMware {code} is a free developer program that provides access to a growing community of developers and experts. If you are an existing Developer Center user or partner program member, you can join the new VMware {code} program using your MyVMware account. If you do not have one, register below to access SDKs and participate in developer forums.

Join VMware {code} with More Than 6752 Developers and VMware Experts

VMware {code} has the fastest-growing developer community devoted to virtualization, cloud computing, DevOps, and other topics.

When you join VMware {code}, you will have the opportunity to connect with VMware experts in the community and participate in outreach programs such as the VMware {code} Speaker Bureau.

Who's On Slack Right Now

For complete message and to post reply visit https://vmwarecode.slack.com/messages/general
2018-12-09
rob.weiss  2018-12-08 14:02:04 CST
@rob.weiss has joined the channel
 
d.vandesande  2018-12-08 06:58:55 CST
@d.vandesande has joined the channel
 
rob.bailey.it  2018-12-08 06:47:51 CST
@rob.bailey.it has joined the channel
 
vmware  2018-12-07 14:53:45 CST
Thanks for your feedback, it is appreciated!
 
stvkpln  2018-12-07 14:48:58 CST
Welp, based on the very limited details you’ve provided, it’s a guess at best. Good luck with that!
 
vmware  2018-12-07 14:02:11 CST
Regarding "idle loops" - KB1077 says: "When an idle virtual machine executes its idle loop, it is actively executing instructions which run on the underlying physical processor. In this case, performance tools in the guest operating system show an inactive operating system, but the CPU on the server is fully utilized." - Which makes sense, but would then expect the physical CPU to show utilized instead of idle.
 
vmware  2018-12-07 13:58:56 CST
Thank you all for your input and suggestions also, I do appreciate the feedback.
 
vmware  2018-12-07 13:58:00 CST
it is relevant to how many vCPUs need to be scheduled. Are you suggesting that I should build a separate cluster for each "size" of VM for equal scheduling? I do get the size of the scheduling slot, I don't understand where the contention is when those resources are idle
 
stvkpln  2018-12-07 13:56:22 CST
Again, aggregated vCPU is only relatively relevant in the context of how large each VM is
 
vmware  2018-12-07 13:35:17 CST
but, there are 64 physical cores in each node in this cluster, no hyperthreading, and the CPU utilization is <5%. There are currently 124 vCPUs deployed on that node, most of which are mostly idle VMs CPU-wise (memory is different however). This node is around a 2:1 vCPU to pCPU ratio which from everything I ready is plenty adequate. The VM in question is trying to schedule 8 vCPUs, but as %RDY indicates, choking.
 

DO MORE WITH MY VMWARE

Ready to dig deeper? In addition to our Slack team, you can register for a My VMware account to download the latest SDKs, get recognition for your code samples, and participate in our online community forums.

Download SDKs

Download the Latest VMware SDKs

Your account gives you access to the latest SDKs for vSphere, Horizon, WorkspaceONE, and other VMware technology platforms.

Forum

Participate in Developer Forums

Share knowledge and interests in the VMware Communities forums with more than 2 million virtualization experts and community members across the globe.

Code

Get Credit for Your Code Samples

Contribute your code samples to VMware Sample Exchange to earn points and get recognition, or request a code sample from the community.

OTHER WAYS TO CONNECT WITH VMWARE {CODE}

Newsletter

Subscribe to the VMware {code} newsletter for access to the latest hands-on labs, blogs, podcasts, how-to videos, and other content delivered directly to your inbox.

VMware {code} Newsletter »

Twitter

Follow VMware {code} on Twitter to get the real-time news and updates from VMware and the community.

Follow @vmwarecode »

GitHub

Explore and contribute to open-source projects such as VMware Integrated OpenStack and the Photon Platform.

VMware on Github »