| Subject | Re: [Firebird-Architect] Transportable databases (3 of 3) |
|---|---|
| Author | Dmitry Yemanov |
| Post date | 2006-11-28T17:38:36Z |
Ann W. Harrison wrote:
1) Am I right that all the swapping code is located in a separate module
that's called from PIO and VIO subsystems? I.e. there shouldn't be any
regressions for native endian databases?
2) Is the "other endianness" DBB flag set only for ODS11 databases? I.e.
is an error still thrown for an ODS10 database that has compatible
alignment but different endianness?
3) Do you have any performance numbers to compare "native endianness" vs
"mixed endianness"? E.g. elapsed time for a backup process run against
two kinds of databases on the same host?
Dmitry
>A few generic questions (sorry if I missed the answers in your posts):
> Essentially these message describe the full implementation of a
> mechanism to allow access to other-endian databases. It seems like
> a useful feature to me.
1) Am I right that all the swapping code is located in a separate module
that's called from PIO and VIO subsystems? I.e. there shouldn't be any
regressions for native endian databases?
2) Is the "other endianness" DBB flag set only for ODS11 databases? I.e.
is an error still thrown for an ODS10 database that has compatible
alignment but different endianness?
3) Do you have any performance numbers to compare "native endianness" vs
"mixed endianness"? E.g. elapsed time for a backup process run against
two kinds of databases on the same host?
Dmitry