Subject | Re: [firebird-support] Database Corruption |
---|---|
Author | Svein Erling Tysvaer |
Post date | 2006-09-02T17:56:41Z |
Could any backup routines interfere with the database? Gbak should be
safe, but file copying isn't. I don't know whether antivirus could
interfere or not. (You don't say whether your database failures happen
randomly or regularly).
Other things that have caused similar problems for people before, is
faulty UDFs and once upon a time there were problems with events (hence
I doubt events would cause problems for you). Are you using any such
thing, or just plain DML? Hopefully, you do not mix DDL and DML within
the same transaction, nor DDL on tables that other people have
uncommitted transactions running against.
If everything fails and the data are important (well, in that case you
should REALLY not use a beta in production), IB LogManager can help you
log changes so that you can rebuild many changes since the last backup
(at least I think so).
It also sounds strange that committed transactions since a certain time
got lost. If you commit, your data should be safe (well, if your
database really got corrupted and you restored a backup from 8:11, then
it would certainly make sense that data inserted later got corrupted.
Set
djolessy wrote:
safe, but file copying isn't. I don't know whether antivirus could
interfere or not. (You don't say whether your database failures happen
randomly or regularly).
Other things that have caused similar problems for people before, is
faulty UDFs and once upon a time there were problems with events (hence
I doubt events would cause problems for you). Are you using any such
thing, or just plain DML? Hopefully, you do not mix DDL and DML within
the same transaction, nor DDL on tables that other people have
uncommitted transactions running against.
If everything fails and the data are important (well, in that case you
should REALLY not use a beta in production), IB LogManager can help you
log changes so that you can rebuild many changes since the last backup
(at least I think so).
It also sounds strange that committed transactions since a certain time
got lost. If you commit, your data should be safe (well, if your
database really got corrupted and you restored a backup from 8:11, then
it would certainly make sense that data inserted later got corrupted.
Set
djolessy wrote:
> Hi all,
>
> Yestardey I had a big problem with database corruption.
>
> Objective:
>
> Fedora 3.0
> Firebird 2.0 RC2
> ForcedWrites = On
> Page Buffers = 2000
> Sweep Intervale = 20000
> 25-30 clients connected by ODBC driver
>
> From time to time database stops (twice a week). Could not find any
> conection with actions of users and database fail. I look in LOG file
> and found 104, 32, 111 errors. After rebooting server everything is
> Ok. Yesterday database faild about 12:00PM, the tehnician reboot the
> server (almost in a same way as power loss). After that database was
> corrupted. All data from 8:11AM was lost and it was a lot of data
> (about 80 invoices).
>
> Where did I go wrong?
> What can I do to stop this behaviour?
>
> Regards Djole