OK - I will use this as an additional opportunity to ask the DB
developer team to implement some kind of journaling as a means for
fast recovery. There is one request for auditing all the SQL
statmenets in the bug-track, but there should be more heavy emphasis
on reliability.

--- In firebird-support@yahoogroups.com, "vincent_kwinsey"
<vincent_kwinsey@...> wrote:

>
> Usually the data pages are get corrupted because they are far more
> than the data pages that keeps metadata, but in this case I guess -
> the datapages for metadata are corrupted. I know one - almost real
> folk story - about such case. Then the developer acquired knowledge
> about ODS (OnDiskStructre) of fdb file and he managed to find and
> correct some byte that caused such errors. Usually there is no more
> than some incorrect bytes. There are some programms that - for at
> lest to Interbase - could display the fdb content as set of records
> and so on, IBSurgeon was one of them. BTW - for C++ programmer this
> can be not so big issue.
>
> --- In firebird-support@yahoogroups.com, "Adam" <s3057043@> wrote:
> >
> > --- In firebird-support@yahoogroups.com, Ahmad $arhan ****
> > <mahdoom_a@> wrote:
> > >
> > > Dear all,
> > >
> > > my problem is getting worst, in addition to the prevoius sql
> error
> > i am getting this when i tried to back-up and restore DB, please
> help me
> > >
> > > select rc.rdb$constraint_name,
> > > i.rdb$field_name
> > > from rdb$relation_constraints rc, rdb$index_segments i
> > > where (i.rdb$index_name = rc.rdb$index_name) and
> > > (rc.rdb$constraint_type = 'PRIMARY KEY')
> > > and (rc.rdb$relation_name = 'RLSTT')
> > > order by i.rdb$field_position
> > >
> > > Unsuccessful execution caused by a system error that precludes
> > > successful execution of subsequent
> > > statements.
> > > internal gds software consistency check (can't continue after
> bugcheck).
> > >
> > > select RDB$RELATION_NAME, RDB$SYSTEM_FLAG
> > > , RDB$DESCRIPTION
> > > from RDB$RELATIONS where (RDB$VIEW_SOURCE is NULL)
> > > and (RDB$SYSTEM_FLAG = 0)
> > > order by RDB$RELATION_NAME
> > > Unsuccessful execution caused by a system error that precludes
> > > successful execution of subsequent
> > > statements.
> > > database file appears corrupt ().
> > > bad checksum.
> > > checksum error on database page 39243.
> >
> > This seems to me to be a simple case of database corruption.
> Database
> > corruption is extremely unusual unless you are mishandling the
> > database files.
> >
> > Gfix may help, or even try a simple backup using the -g switch.

If

> the
> > corruption is in a page that contains no useful data, then you

may

> be
> > lucky and can take a working backup by avoiding the normal garbage
> > collection during the backup.
> >
> > Adam
> >
>