Well, I'll drop my 2 cents into the fray. As a DBA, when a database = failure occurs I want to be able to recover the application/database as = applicable to the point in time that "all hell broke loose". Personally = I don't really care about the "Business needs" as they change faster = than most runway models. More importantly, given that there is a stated = business requirement and action plan when the application fails for a = database error those in power will not be reading the action plan, but = reacting to pressures from higher up who never signed off on that action = plan in the first place. Therefore, DBA beware. Failures are = inevitable, it's just a matter of when, how bad, and how much "help" = you'll have breathing down your neck!! Dick Goulet Senior Oracle DBA Oracle Certified 8i DBA -----Original Message----- From: Daniel Fink [mailto:Daniel.Fink@xxxxxxx] Sent: Monday, May 03, 2004 3:42 PM To: oracle-l@xxxxxxxxxxxxx Subject: Re: Oracle recovery I recall a paper from Oracle (about 5 years ago) that analyzed the various outages, the methods of recovery and how long they took. I have it at home and I'll see if I can get a doc_id from it. The point I would make with the DBA is that exports are not a recovery mechanism, they are only able to restore to a given point (time of the export) and they do not restore everything (sys objects). With an export, there is no way to roll-forward. Of course, a cold backup w/out archive logs is also unable to roll-forward. All data from the last export and the current time is lost period end of story. Is this acceptable? Perhaps. In certain environments where data loss is not a problem, exports are a pretty simple way of capturing a snapshot of the data. It is also not just 'data' that is lost. It is anything stored inside the db, including procedures, packages, etc. In a development environment, a days worth of coding could easily run into the 10s of 1000s of dollars. Daniel "M.Godlewski" wrote: >=20 > Believe it or not the database I'm currently using the DBA is forcing = us to use exports as our applications recovery scenario. I'm trying to = locate information about archive logging and the tablespace point in = time recovery versus full database recovery. I wanted to find some kind = of percentages about the number of databases that need full database = recovery (ie applying all archive logs) and percentages of databases = that only needed TSPIR. >=20 > I thinking if I can get some type of failure rate it could help sway = him into using archive log mode which would give us better recovery = options. ---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request@xxxxxxxxxxxxx put 'unsubscribe' in the subject line. -- Archives are at //www.freelists.org/archives/oracle-l/ FAQ is at //www.freelists.org/help/fom-serve/cache/1.html ----------------------------------------------------------------- ---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request@xxxxxxxxxxxxx put 'unsubscribe' in the subject line. -- Archives are at //www.freelists.org/archives/oracle-l/ FAQ is at //www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------