Hi Alexander, That did not work. I tried issuing DELETE using curl as you suggested, but server always responded with not_found error. Is there any way to clean replication state manually on the file system?
Andrey On Tue, Aug 26, 2014 at 12:06 PM, Alexander Shorin <[email protected]> wrote: > On Tue, Aug 26, 2014 at 7:50 AM, Andrey Kuprianov > <[email protected]> wrote: > > So the question is: how does one fully clear the > > replication state from one db to another, so I can start replication > from A > > to B from scratch? > > In your logs find the line which looks like this one: > [Mon, 25 Aug 2014 04:45:32 GMT] [info] [<0.28499.6>] starting new > replication `4f425b4674543655d4486ad7ca9c9bc2+continuous` at > <0.28505.6> > > The `4f425b4674543655d4486ad7ca9c9bc2+continuous` is your replication > ID. Now, to delete the replication history just run two requests for A > and B databases: > > DELETE http://server1/A/_local/4f425b4674543655d4486ad7ca9c9bc2+continuous > DELETE http://server2/B/_local/4f425b4674543655d4486ad7ca9c9bc2+continuous > > And also make sure that replication isn't active at this time since > replication process hold the state. These _local docs are that state > snapshots for recovery. > > -- > ,,,^..^,,, >
