Data Backup and Recovery
Data Backup and Recovery
Hi Guys,
I was invited to ask a couple of question, and what do you know, I have a meeting on Wedn. to meet the SQL/Sharepoint admins to discuss how to design a migration
over to our NTAP controllers. I have in the past participated with design but never for these products so I feel a bit lost, but not too much 😉
I get the drift where one needs to segragage transient data from non-transient data in different volumes and other basic stuff. What I d like to ask, is there a good doco. that explains
1. basic rules with respect to designing this and how to do capacity planning?
2. what protocol or options to protocols have I got? is it CIFS only or can FC work as well? Im not a big fan of CIFS for apps.. its way too chatty.
3. what other info should I get? Should I drill down to read/write patterns and use UDP or TCP or does that even come into play (only applies for CIFS I know).
Thanks,
eric
Solved! See The Solution
Eric,
Planning the SharePoint/SQL environment you need to keep in mind the number of users, the expected online number of concurrent users and the expected data size per user. This will help you factor in the capacity that you will need for your SharePoint farm. Ensure that you are planning for capacity for your SharePoint environment. FC and ISCSI LUNs and volumes are recommended over CIFS. Ensure that in the SQL Server you separate out the tempdb and system databases on separate volumes that your data and log. Tempdb you do not have to take snapshots of since it is all transient data and is rebuilt ever time you restart the SQL instance. And the backup for your system databases is a streaming backup other than a snapshot. If you need more about specific sizing requirements please contact your NetApp representative and have them go through a sizing exercise with you.
John
Have you seen these yet?
http://media.netapp.com/documents/esg-wp-netapp-sharepoint-server.pdf
http://media.netapp.com/documents/tr-3675.pdf
Ours web tier running on a VM over FC and the SQL server is a phyical MS cluster, again FC.
Bren
Without going into depth (right now at least ), the biggest thing to me is understanding how the "point in time" restore capability affects sizing -- particularly around the "snapinfo" directory.
If you do go with "point in time", you can use a startling amount of space (barring thin provisioning...but that's a whole different discussion ).
Thanks guys,
Brendon, I remember you wrote a doc. about this and I was going to ask you for it, so thanks for that!
allright, back to my p2 issue.
Eric
Eric,
Planning the SharePoint/SQL environment you need to keep in mind the number of users, the expected online number of concurrent users and the expected data size per user. This will help you factor in the capacity that you will need for your SharePoint farm. Ensure that you are planning for capacity for your SharePoint environment. FC and ISCSI LUNs and volumes are recommended over CIFS. Ensure that in the SQL Server you separate out the tempdb and system databases on separate volumes that your data and log. Tempdb you do not have to take snapshots of since it is all transient data and is rebuilt ever time you restart the SQL instance. And the backup for your system databases is a streaming backup other than a snapshot. If you need more about specific sizing requirements please contact your NetApp representative and have them go through a sizing exercise with you.
John
Hi John,
thanks for a bunch, that ll get me going! I ll mention NetApp as a service provider to my manager as well, no worries.
Cheers,
Eric