06-10-2021 12:10 PM
Please... can anyone explain why, it costs nearly $100k in hardware now just to run MSO? Now that any future version of MSO is part of Nexus Dashboard, it took the already crazy oversized requirements of MSO and made it worse.
Nexus Dashboard
For 2.0.2h it requires either a physical cluster of 3x SE-NODE2-G2, which MSRP at ~$35k each. ($105k + support)
Or a cloud option of 3x 16 vCPU, 64Gb ram, 350GB storage. For a cloud spend of ~$20k / year.
MSO shouldn't cost this much to run...
06-10-2021 01:33 PM
Virtual ND is coming out next couple weeks. That should solve your concerns.
Robert
06-11-2021 01:31 PM
It's out, 2.0.2h. It's just hot garbage for vmware.
https://software.cisco.com/download/home/286327743/type/286328258/release/2.0(2h)?i=!pp
1. ova manifest has a sha256 mismatch.
2. The deployment script for it errors out on the data drive because it left an open file reference for an image that returns "null" and bombs.
3. Still requires 16 vcpu and 64 gb ram per node. So hardware cost is still extreme.
4. Its running on Centos 7 which is a dead end OS, furthermore the ova has vmxnet3 network adapters selected which don't work.
5. vmware tools aren't installed on the image so all of the customization of the vm never occurs from the template.
6. no root access to the vm to fix any of the above centos level issues. Only account is rescue-user which can't do much aside from grab the token.
7. ACS command line is poorly documented so much so that even --help of the commands don't return anything.
8. The token used for nodes 2 and 3 is time sensitive, as it expires in 30 minutes. So you not only have to rush to get nodes 2 and 3 up, but if you don't there isn't a documented cli remediation to use a new one and have to destroy the vm and go back to the template deployment.
Lots of work needed...
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide