The slave02 server will be a copy of the DB into Amazon. There will be a migration to Amazon in the future, and the company's managers want a copy of the current DB there to test and do all the stuffs they need (migrating to 9.5, too).
slave01 is already working as a streaming replication server.
The master server sends the wal_files to slave01 by archive_command.
The plan below isn't my idea, I would do different but isn't my call:
Current scenario:
master stores wal_files into slave01
slave02 does not exists
The plan is:
1. setup slave02 at amazon EC2 (just for testing and future master server for devs)
2. setup postgres on slave02 (9.2)
3. pg_basebackup will be run from slave01. This will split the base in files of 50GB each (example)
4. Send the splitted files from slave01 to slave02
5. restore/join the files
6. start postgres on the slave02 slave
7. restore the DB using the wal_files from slave01
Question:
Is possible to make slave01 archive the wal_files? can just be a archive_command and archive_mode = on?
