Subject | Re: GBAK size limit question (again, sorry), also BLOB questions |
---|---|
Author | jeeggers |
Post date | 2003-09-22T15:25:52Z |
--- In firebird-support@yahoogroups.com, "Artur Anjos" <artur@a...>
wrote:
itself".
Ooops, I meant "backs up the database itself"... But you answered my
question on that, thanks.
mechanism to allow the front-end to access the blobs from a central
location. I'm pretty sure I want to store the blob in the database
but I'm not sure if it makes sense to store it in the same table as
the other data for that row, or if I should have a separate table
that contains just the blob and a blob-id, then link the row in the
blob-table to the data-table through that blob-id. My reasoning for
that would be that if the non-blob data is updated, it doesn't create
a whole new version of the blob data.
wrote:
> only does what you told him to do. Gbak creates a backup file, Idon't understand what you are talking to with "updates the database
itself".
Ooops, I meant "backs up the database itself"... But you answered my
question on that, thanks.
> > Another question, about BLOBs: I'm storing BLOBs as big as 120 MBYeah, but I'm thinking that would require a 2nd client/server
> > per row. Am I insane ?
> Some people will tell you that it will be ok, other's the oposite.
> There is no tech problem to have such a BLOB, the argument to keep
> the blob outside the database and just keep a pointer to it inside
> it's very simple: Blob data with that size don't change too much,
> so it will be easy to backup this kind of database if you keep it
> outside the database. Get it?
mechanism to allow the front-end to access the blobs from a central
location. I'm pretty sure I want to store the blob in the database
but I'm not sure if it makes sense to store it in the same table as
the other data for that row, or if I should have a separate table
that contains just the blob and a blob-id, then link the row in the
blob-table to the data-table through that blob-id. My reasoning for
that would be that if the non-blob data is updated, it doesn't create
a whole new version of the blob data.