Re: Data Guard (zero data loss)

In article <6de2ae57.0303100354.15b5658f_at_posting.google.com>, piet.de.visser_at_cmg.com says...
>

snip
>
>Tim, Group,
>
>This question has come across my desk at least 5 times this month (is
>it the season ?).
>
>May I join Tim in asking :
>Does anyone have REAL experience (e.g. serious test or production)
>with DataGuard and ZERO (or very-near-zero) loss of data ?
>

Yes.

>Did you have the absolute NEED to ?

Yes.

>If-so, was it worth it ?

Yes

>
>By lack of real references, I keep recommending to just send (ftp) the
>arc-logs
>to a safe location, and apply them regularly, or ad-hoc.
>
>If you compress (gzip) your arcs before sending, it also consumes a
>lot less network capacity, but this requires the balls to trust the
>compression software
>and the script that does the g(un)zip.
>
>Regards,
>
>PdV
>Oracle DBA

If that works for you fine. I'm always surprised by people feeling the need to do MORE than they have to. Are there many sites that need this sort of disaster recovery capability? No, but they do exist. When I first saw this stuff in the 9i beta, my reaction was pretty much the same as yours - "Why the heck would anyone need that?" But somebody must have - otherwise it wouldn't be in the product. Why waste development lifecycles (of which there are never enough) building specific functionality that is never going to be used? Note here that I'm talking about functionality that is an enhancement to existing stuff, rather than "let's make a new product because we think it's wonderful and everyone will want to go out and buy it" crappy marketing decisions.

In this particular area, I know of one customer in Europe who has implemented this feature, did need it and are happy with it. Are there others? Probably. I don't deal with every customer in the world! :)

HTH. Additions and corrections welcome.

Pete

SELECT standard_disclaimer, witty_remark FROM company_requirements; Received on Mon Mar 10 2003 - 12:07:55 CST