I do this each hour for backup purpose.
The behavior is as you describe - the replicated DB is bigger than the
source.
After replicating with curl, I do a compact on the replica, and then its
size is reduced.
------------------------------------------------------------------------
*With best regards,*
Kiril Stankov
On 10-Oct-15 8:43 AM, Mike Marino wrote:
Did you try compacting the target database?
Dave Westerman <[email protected]> schrieb am Fr., 9. Okt. 2015
um 23:55:
Why would a new target replica of a CouchDB database be over twice the
size in the filesystem as the source database after the replication is
finished? I have a CouchDB database that has 2.5 million documents,
and takes up about 597MB on the filesystem. When I replicate to
another database on the same system, the size of the replica is 1.4GB.
I have deleted and recreated both the source and target databases from
scratch, and populated the source. So the replica is totally new, and
doesn't have any other data in it.
