What now for VMWare? What now for virtualisation?

I've spent months testing performance from esxi to kvm and single core and disk performance is a good chunk worse and I have tried everything, it's a a shame. I tried the same workload on aws though and it's even worse than local kvm so there is that.
 
So very true. And further to that, when you want to test BC/DR plans, oh no we can't possibly do that. Then wonder why when you need to invoke it everything falls on its face.
"We need to test our BCP at least every quarter. We've not tested for over a year now..."
"We can't afford the downtime or paying the staff overtime to test that."
*Experience a catastrophic outage, enacting the BCP*
"WHY IS NOTHING WORKING???"
"Because you didn't want to test it regularly, and now we're having to fix things in tandem with enacting the BCP."
*Company loses 10x what the testing would've cost."
 
How are you getting on with it?
I Had a meeting but their kvm is juat branded no chanhlhe though i have to test it later in the year for the company. Most of my time had been tesring nearly every option on the libvirt xml to test performance. Found a few bits to improve and found that with kvm and hyperthreading things like passmark benchmark show much worse thsn vmware with hyperthreading on.
 
We're currently moving our infrastructure from on-prem vmware to Azure with the rest of our stuff. Can't wait to get off it, Broadcom can suck a lemon.

Doing this also. Got my f u email a little while ago. It was expected but it sucks given how much we have invested both in cash and knowledge in VMWare
 
Doing this also. Got my f u email a little while ago. It was expected but it sucks given how much we have invested both in cash and knowledge in VMWare

Poor @ChrisD. is getting twitches with that captial W in VMware :D

Funnily enough had a VMware day today. Troubleshooting the old NSX-T native Load Balancers and they ABSOLUTELY SUCK!
 
Poor @ChrisD. is getting twitches with that captial W in VMware :D

Funnily enough had a VMware day today. Troubleshooting the old NSX-T native Load Balancers and they ABSOLUTELY SUCK!

They can capitalise **** VMWARE

Not sure if the plan is working, or is it just driving customers elsewhere? We can't justify reupping with Broadcom for what we do, it isn't worth the money
 
They can capitalise **** VMWARE

Not sure if the plan is working, or is it just driving customers elsewhere? We can't justify reupping with Broadcom for what we do, it isn't worth the money

We're the opposite, we can't justify cloud-native with what we do. We're formerly mostly AWS and have gone all in on VCF. A re-architecture and shift in development process would reduce hyperscaler costings, but that shift isn't happening for a long, long time. Plus, I've said many, many times - I don't believe any other SDN comes close with it's agentless, powerful capabilities than NSX, and I'm someone who has basically been forced into NSX roles for the last 6-7 years and not a huge fan of working with it.

I do feel for small SMBs though, they've been hung out to dry because they're simply not BC target audience.
 
I Had a meeting but their kvm is juat branded no chanhlhe though i have to test it later in the year for the company. Most of my time had been tesring nearly every option on the libvirt xml to test performance. Found a few bits to improve and found that with kvm and hyperthreading things like passmark benchmark show much worse thsn vmware with hyperthreading on.
I know they've discontinued it now - but I saw better performance using Clear Linux (the intel maintained and optimised kernel).

So if any distro has picked up that kernel or it's patches then it might be worth a look.
 
Back
Top Bottom