Re: Corrupt redo log file - database only gets to mount

Like I said, the redo logs were not duplexed. Three redo log groups with only a single member in each. Very bad for recovery as they've now discovered. If they were duplexed (as best I can recall), it would be a simple matter of copying one of the good member files in place of the bad one. (I'm sure someone will correct me if I'm wrong on that point. :)

As it is, I was thinking if they have a good copy of that redo file on tape and since the database is in archivelog mode, then maybe they could get the system back up that way. That may not help as you suggest, but I was grasping at straws to avoid having to tell them for sure they were looking at a rebuild because their system was poorly laid out and managed.

If I actually get tasked with helping them beyond what I already discovered, then I'll be calling Oracle. Just thought I'd post tonight here to satisfy my curiosity and see if there was something obvious I could direct them to do... other than call Oracle, of course. :)

It's a tiny database and I'm guessing is not production critical since they waited around to get me on the phone and didn't seem panicked about it being down.

On 5/13/02 7:31 PM, in article 3CE04CD8.8460AB6F_at_exesolutions.com, "Daniel Morgan" <dmorgan_at_exesolutions.com> wrote:

> Jeff Traigle wrote:
>
> I'm a little uncertain of what you mean by back-up copy of the redo log. There
> are hopefully multiple redo logs in each redo log group and multiple groups.
> But
> these, alone, are not going to get you very far. Especially if they were
> backed
> up as the backup likely bears no relationship to the state of the log files at
> the moment they became toast.
>
> Can you clarify?
>
> I think, were I you, I would have opened an iTAR at metalink by now. This has
> all the signs of being ugly.
>
> Daniel Morgan
>

-- 
Jeff Traigle
traigle_at_si.umich.edu
http://www-personal.si.umich.edu/~traigle/