Hi everyone,
we're interested in taking advantage of batch replication though we have troubles configuring it the right way.
On publish instance we've set up Flush Agent to support batch mode. Also HTTP Method was changed from GET to POST to allow list of paths passed in request's body.
![screenshot.png screenshot.png](https://experienceleaguecommunities.adobe.com/t5/image/serverpage/image-id/22290iA53B4FD7DE0B9896/image-size/large/is-moderation-mode/true?v=v2&px=999)
Then when Flush Agent performs the request it returns 200-OK response. However there are errors in dispatcher log:
[Sun Feb 02 23:02:40 2020] [I] [pid 12 (tid 140047550482176)] Activation detected: action=Activate [batch:]
[Sun Feb 02 23:02:40 2020] [E] [pid 12 (tid 140047550482176)] URI not canonical: batch:.
[Sun Feb 02 23:02:40 2020] [E] [pid 12 (tid 140047550482176)] URI not canonical: batch:.
[Sun Feb 02 23:02:44 2020] [I] [pid 12 (tid 140047550482176)] "POST /dispatcher/invalidate.cache" 200 13 4560ms [toryburch_farm/-]
On top of that stat files created in directories except the ones where they should be therefore auto-invalidation fails.
Could someone please give us an advice on how to configure the agent properly?