To load a transaction log into an archive database, use:
load tran db_name from dump_device [ [stripe on stripe_device] ... ]
where:
db_name – is the archive database into which you are loading the transaction log.
dump_device – is a local disk transaction log dump.
stripe_device – specifies the stripes for the local disk transaction log dump.
When you load a transaction log into an archive database, load tran runs the recovery redo pass. Modified and new database pages are written to the permanent changes segment. You must have enough space in the modified pages section to accommodate these changes. If necessary, increase space for the modified pages section by using alter database to increase the normal database storage allocated to the archive database.
Unlike a traditional database, an archive database can be brought online in the middle of a load sequence without breaking the load sequence. When a traditional database is loaded and then brought online without using the for standby_access clause, it is no longer possible to load the next transaction log in the load sequence. An archive database however, can be brought online without the for standby_access clause and later, loaded with the next transaction log in the load sequence. This allows read-only operations like running consistency checks, at any time during the load sequence. This is possible because when loading a transaction log into the archive database, Adaptive Server automatically removes the disposable changes segment from the modified pages section. This effectively reverts the archive database to its state after the previous load was done, thereby allowing the next transaction log in the sequence to be loaded.
