I made another backup, but this time I split the backup into mulitple
3GB files. This backup restores properly and results in an
operational database.

Just to be sure, I tried to restore the single file backup that was
automatically made last night. The restore failed in the same place
all the previous ones had - right around the time the resulting
database hits 4GB.

So: have I found a bug? If so, is it in GBAK or in the engine?

--- In ib-support@yahoogroups.com, "jeffshoaf" <jeffshoaf@a...> wrote:
> We have a Firebird 1.0 database that has exceeded 6GB. Since the
> server is running Windows 2000 with NTFS as the file system, we're
> trying to restore a backup made with GBAK to a single file using
the
> Localhost connection.
>
> Whenever the file gets to 4GB, the server shuts itself down and the
> restore fails with a lost connection. The server log lists the
> following problems with the same timestamp:
>
> INET/inet_error: accept errno - 10093
> INET/select_wait: select failed, errno = 10093
> SRVR_multi_thread/RECEIVE: error on main_port, shutting down
>
> Then, 2 seconds later:
>
> C:\Program Files\Firebird\bin\ibserver.exe: terminated abnormally (-
1)
>
> The guardian then restarts the server.
>
> This always happens at the 15,700,000th record and the
resulting .gdb
> database file is exactly 4GB in size.
>
> Can't a database that's larger than 4GB be restored into a single
> file database?