||One source of downtime often experienced relates to loss of storage media. ASM can meet this challenge in, essentially, the same way that it can be met by other technology solutions such as RAID.
But ASM has an additional capability that no other technology can address: The ability to perform a seamless, zero downtime, migration from one vendor's SAN to another.
You can move your application, while in use, between SAN, NAS, DAS, iSCSI, and other storage technologies while the blocks are being accessed. Bet EMC will never help you move off your Symmetrix to a NetApp.
||Audit Vault may be sold by Oracle primarily as an audit enhancement but full disclosure requires that one acknowledge that if you have an audit of every transaction ... you have an enhanced ability,
through Flashback ... to fix substantial damage without downtime.
||Data Guard comes in three flavours ... physical, logical, and snapshot. For high availability in cases where there is the loss of an entire data center,
for example during a natural disaster or some moron used a backhoe to tear up more than just the street, physical Data Guard is the best solution available.
||One downtime event that was unavoidable until Database 11gR2 was downtime during application upgrade when upgrading synonyms, views, and PL/SQL objects.
Editioning makes it possible to eliminate this source of downtime and Edition Based Redefinition has been further extended since its intial release.
||Flashback database is, to paraphrase Oracle, an ability to rewind the database as though it was a movie to a previous point-in-time.
||Flashback drop uses recyclebin to allow the complete recovery of dropped tables with their constraints, indexes, and triggers.
||Flashback table can be used to take a single table back to a previous point-in-time without affecting any other database operations.
||The built-in DBMS_REDEFINITION package makes possible the real-time alteration of tables, adding columns, dropping columns, renaming columns, manipulating data,
and at the same time changing constraints and triggers.
|Real Application Clusters
||RAC is intended as an in-data center solution to loss of an instance or server. It eliminates dependency on any single piece of hardware at a single, or closely spaced, locations.
||RMAN is far more than a simple tool for backup and recovery: Ask anyone that has done a real-time, zero down-time, block level recovery: It can be a life-saving experience
||When transactions run out of space due to an inability to enlarge a segment or tablespace things stop ... and rollback.
Resumable Transactions are transactions that suspend themselves and restart automatically when space is available.
|Server Parameter File
||Oracle has some fancy market-speak name for this but the simple fact is that a tremendous amount of reconfiguration can
be done with zero downtime by properly using an SPFILE and understanding how to properly SCOPE your ALTER SYSTEM statements.
||Transaction back-out was added to the DBMS_FLASHBACK package with Database 11gR1. This capability allows the backing out of a committed transaction while the database is open and in use.