Hi All,
We're about to implement on 6.4 in our new datacenter where we currently have about 350 VM's. In the next year of so this number is going to shoot up as we migrate VM's from three other datacenters. So by the end of the year I'd expect there to be at least 1000 VM's. We currently have 12 large Esxi hosts (512Gb RAM, quad processor 10 cores per processor) and more will be added when performance needs as VM's are migrated from our old datacenters ..all production VM's are replicated by SRM to a recovery datacenter.
In addition the powers that be are about to release a 'self provisioning' tool that will allow end users to deploy as many VM's as they desire from a web based front end..(Nice!)
All backups are to disk which is replicated to the recovery datacenter.
Our current TSM retentions are fairly strict with 3 version over the last 50 days. I'm going to suggest we use the same retention with an incremental forever backup.
re data movers and how to effectively schedule backups. I thinking at least one per ESXi sever with two schedulers ? Maybe one scheduler can backup production VM's the other Test/Dev/UAT ?
I understand that each scheduler can backup more that one VM simultaneously (that correct ?) I'm not sure what is a sensible number to start at here (5?).
I also have concerns about how the operations team can quickly discover backup failures and take appropriate actions..
Any thoughts (an no doubt more questions) would be greatly appreciated...
We're about to implement on 6.4 in our new datacenter where we currently have about 350 VM's. In the next year of so this number is going to shoot up as we migrate VM's from three other datacenters. So by the end of the year I'd expect there to be at least 1000 VM's. We currently have 12 large Esxi hosts (512Gb RAM, quad processor 10 cores per processor) and more will be added when performance needs as VM's are migrated from our old datacenters ..all production VM's are replicated by SRM to a recovery datacenter.
In addition the powers that be are about to release a 'self provisioning' tool that will allow end users to deploy as many VM's as they desire from a web based front end..(Nice!)
All backups are to disk which is replicated to the recovery datacenter.
Our current TSM retentions are fairly strict with 3 version over the last 50 days. I'm going to suggest we use the same retention with an incremental forever backup.
re data movers and how to effectively schedule backups. I thinking at least one per ESXi sever with two schedulers ? Maybe one scheduler can backup production VM's the other Test/Dev/UAT ?
I understand that each scheduler can backup more that one VM simultaneously (that correct ?) I'm not sure what is a sensible number to start at here (5?).
I also have concerns about how the operations team can quickly discover backup failures and take appropriate actions..
Any thoughts (an no doubt more questions) would be greatly appreciated...
