Re: Datawarehouse backup

> As Scott mentioned, the dw database generates huge amount of
> archivelogs The reason we are not running in archive log mode is that
> the backup servers can't keep up with. That is why we decided to go
> with NOARCHIVELOG mode.

Somewhat reminiscent of Donald Rumsfeld's comment that "you go to war with the army you have". While it depends on the decisions and attitudes made by those in charge, one has to ask oneself if those decision-makers were provided with the right information by those who implement.

(The analogy, of course, breaks down somewhat when one considers the constraints imposed by military discipline and the UCMJ, not present in corporate environments. But the politics and the personal dynamics are very similar.)

Were decision-makers made aware of the risks of the present state of affairs (i.e. "backup subsystem insufficient" or "ETL processes need to be tuned")? Or, was this alternate solution presented without mention that the goalposts were moved as well?

To put it another way, are the decision-makers (who have purchasing authority for a more robust backup system) aware of the *probability* of multi-day downtimes due to multi-Tb restores from an inadequate backup subsystem and more than a week's worth of ETL "catchup", which are not guaranteed to recover the database to the same state as before?

Best of luck. Sincerely.

--
http://www.freelists.org/webpage/oracle-l