Re: Data recovery - URGENT

Поиск
Список
Период
Сортировка
От Jan Wieck
Тема Re: Data recovery - URGENT
Дата
Msg-id 3EED0B86.3000908@Yahoo.com
обсуждение исходный текст
Ответ на Data recovery - URGENT  (Peter Galantha <argosz@tensa.net>)
Список pgsql-hackers
Peter Galantha wrote:
> Hello,
> 
> is there a way to recover data from a damaged database if 
> a.) we have some but not all datafiles
> b.) we have all datafiles but system tables are currupted
> 
> The situation is extremely critical so we're interested in all possible 
> solutions even if it's time consuming or if we need to pay for it!

Okay, data in lost datafiles will be lost (surprise, surprise).

But for the other parts ...

Corrupt or missing index files are no big deal (as long as they are user 
indexes.

Corrupt system index files ... hmmm ... don't remember if v7.1 did a 
good job on system index recovery. Worst case we might have to consider 
that a corrupt catalog.

If system tables are corrupted, the complete schema of the database must 
be known otherwise. If not, this will become some detective work to 
figure out the table structures. Maybe another, older backup exists and 
the schema didn't change after that.

The real key is the commitlog. If that is available, the data files that 
can be recovered could be injected into some sort of emergency room 
database system at least.

I am already constructing a v7.1 testbed, and will be able to answer 
more detailed questions in a few minutes. For speed you can reach me on 
+1 978-697-7733.


Jan

> 
> Version: Postgres v7.1
> OS: Debian Linux 3.0
> Storage: RAID 5 and RAID 1 (the array is dead and we might not be able to 
> restore all data, but there are good chances to recover about 90% of the 
> data)
> 
> Thanks & regards,
> Peter Galantha
> argosz@tensa.net
> 
> 
> ---------------------------(end of broadcast)---------------------------
> TIP 3: if posting/reading through Usenet, please send an appropriate
>       subscribe-nomail command to majordomo@postgresql.org so that your
>       message can get through to the mailing list cleanly



-- 
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#================================================== JanWieck@Yahoo.com #



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Markus Bertheau
Дата:
Сообщение: Re: [GENERAL] UTF8 and KOI8 mini-howto
Следующее
От: "Christopher Kings-Lynne"
Дата:
Сообщение: Re: security flaw