throughput decrease when using more vcpus about osv

[avikivity]

Also the allocator is single threaded.
On Jun 22, 2013 10:54 AM, "loudinthecloud" [email protected] wrote:

A locking issue that cause a lot of context switching, probably due to the
single thread design of callouts dispatching, or assumptions of the freebsd
TCP stack.

• Investigation of the real cause

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/5
.

[loudinthecloud]

Right... and it works hard during netperf.

I'll first work on finding the cause for the context switches and try to understand our CPU utilization in general.

@avikivity, is the allocator something you're going to handle? or would you like me to do it?

[avikivity]

I'd like too but let's plan it.
On Jun 22, 2013 11:27 AM, "loudinthecloud" [email protected] wrote:

Right... and it works hard during netperf.

I'll first work on finding the cause for the context switches and try to
understand our CPU utilization in general.

@avikivity https://github.com/avikivity, is the allocator something
you're going to handle? or would you like me to do it?

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/5#issuecomment-19853284
.

[nyh]

Can someone please close this issue? It's outdated - our memory allocator, callouts and TCP stack are all designed now for scalability. We probably still have multi-core issues, but they are not described in this issue.