Hello,
Does anyone know if it is possible to use SAN Splitting (the function of
splitting a mirror of disks so that there are two idential copies of a
Postgres Instance)? The reason that I ask is because I'm working on a
Hashing database to scale postgres to the moon and I'd like to be able to
create exact replica's of my Postgres database and then divide the
throughput of the data independently to both. (through a hashing algoithm,
I'd be able to then split the I/O to each of the nodes evenly...)
The question basically boils down to whether I can have two idential copies
of the DB running on different servers at the same time. I guess as long as
I have the Server Name be different on each server, the database platform
should not care... Once the split was done, they would go on their own
independent way and take a new life of their own. I would then delete 1/2
of the data from each node. (opposite data of course so that it's still
logically one unit) and I'd be able to provide twice the throughput because
of this.
Thoughts?
--
View this message in context: http://www.nabble.com/Using-SAN-Splits-to-instantly-copy-a-DB-tf2632763.html#a7348176
Sent from the PostgreSQL - general mailing list archive at Nabble.com.