[Info-vax] Real Usenet clients, was: Re: backups and compaction or nocompaction might be better

[pcoviello at gmail.com]

On Tuesday, January 29, 2013 11:23:26 AM UTC-5, pcov... at gmail.com wrote:
> On Tuesday, January 29, 2013 10:47:14 AM UTC-5, Stephen Hoffman wrote:
>
- show quoted text -
Paul,

A "freeze" of the application is likely not enough. The database must be completely inactive to get a valid copy.

As Hoff noted, one possibility is excluding the database-related items from the BACKUP, and using the database's own tools to produce a presumably consistent backup of the database.

If a full BACKUP image is needed, one can use volume shadowing to your advantage to minimize downtime: add a member to the shadow set, wait for the copy to be fully synchronized, quiesce the database, disconnect the EXTRA member, release the database back into production.

The removed shadow set member is a current copy of the (momentarily idled) database. It can then be mounted privately and a BACKUP/IMAGE can be done.

- Bob Gezelter, http://www.rlgsc.com 


lets see if this one posts...

anyways the freeze is at the database level and are actual commands from 
intersystems,  I know for a fact they work, for years we have been doing this 
and on a weekly and as needed basis refreshed our test and training systems 
with the snapclone data 

regards 
Paul