Enabling Storage Optimization Features in New Databases
Daniel Nashed – 9 November 2009 16:38:56
When you have enabled storage optimization features like LZ1, data+design compression and DAOS, you want new databases and replicas also to have the same storage optimization options turned on when the database instance is created.
I have tested with ND 8.5.1 different ways to create new databases and most settings are preserved even when creating a new mail-file from template or when a new archive is created from the mail database.
The changes are propagated thru the flags in the icon note which are also copied when you create an archive or a new database from template.
So if you want new databases automatically have the right storage optimization features ensure that the right settings are enabled on the template already.
This is great news and helps to ensure that the settings are automatically set for new databases.
In the migrations phase you still have to enable those settings manually on each instance of the database (because the flags are only propagated when a new instance is created).
While testing I ran into one issue which is currently a bigger road block for us and worked before in 8.5.
When the archive is created the DAOS status of an archive database is not preserved when the compact -a creates a new archive.
All other settings are correctly preserved only the DAOS flag is missing.
I have opened a new PMR and got a "SPR # SHEZ7XGL5H2 DAOS flag is not preserved when creating a server based archive via compact -a"
According to support I am the first customer reporting this regression. We need this functionality in a larger migration where we push around 2 TB of data from a single mail-server to an archive server and due to storage constrains we need the NLOs directly in the DAOS store which is impossible for us to enable in time during the archive operation thru this issue.
If this functionality is important for you or your customers, you should open a PMR to add weight to this SPR to get it fixed in the next FP.
-- Daniel
- Comments [0]