This content has been marked as final. Show 3 replies
For question one I would refer to this MOS document :
How To Calculate The Required Network Bandwidth Transfer Of Archivelogs In Dataguard Environments [ID 736755.1]
For question two
I'm thinking its not common at all. First question here I remember seeing it. This is the closest and only MOS doc I can find :
Sun ZFS Storage Appliance SW 2010.Q3.4 Release and Product Update [ID 1343086.1]
DR site for an Oracle based product which is hosted in a virtual environment (VMWare ESX).I think you should check with Oracle first, they do not support Oracle running on VMWare. They only support,if it is running Oracle provided VM.
1. What I would be interested to know is what are the typically minimum requirements for networking the two sites if DG is to be used.The site can be any distance apart.
2. Can you tell me, is it common to use a mix of Dataguard and VMWare Site Recovery Manager?When you say, "mix of DG & VMWare Site Recovery Manager" did you mean to club both technology together to implement DG?
From what I know of the two products it seems reasonable to build a DR site with virtual machines using SRM and DG for ensuring the DB is available to those two sites?
I don't think you will be able to do that as DG automatically manages the transactions and replication.
Thanks for your input.
I looked on the VMWare site http://www.vmware.com/solutions/partners/alliances/oracle-vmware-support.html
"Oracle has maintained an official support policy for VMware for over four years. For customers with long-standing Oracle & VMware relationships, this is well understood and proven."
I have a feeling the client is running the Oracle DB on Oracle Unbreakable Linux - which could of course be run in VMWare.
I guess I will have to bug the Oracle DBA's - I don't like to because they are so stretched already :)
They are already using DG successfully with SAP apps. In this case it is Documentum and it's fairly typical that a lot of Ora DBA's are not familiar with EMC/Documentum so much - which is fair enough of course but I know Documentum and a bits or Oracle/Unix and its a case of pulling together all the disciples to get a good HA/DR solution for a Documentum implementation.
When you say, "mix of DG & VMWare Site Recovery Manager" did you mean to club both technology together to implement DG?Yes. I only deal with stuff on top of the DB layer as a rule so I am in a territory I am not familiar with and thus not be making myself well understood.
The system we have is 14 virtual machines including a weblogic layer, content server layer (EMC Documentum) which has a managed file system and behind that is the Oracle database layer (which is a bit of a black box on my Visio diagrams - I know it is RAC).
The file system exists on external storage mounted on VM's running RedHat. I "assume" the Oracle dbf files are almost certainly mounted on some other high performance storage.
In the case of the file system and Oracle data files then they would be external to the application/system VM's. So I was thinking SRM would take care of the VM side of things and DG would take care of the Oracle DB.
We (us Documentum consultants) have used rsync to keep the file-content (the bit managed by Documentum) sync'd between live and standby sites when other tools are not available.
Many thanks for input from you guys.