Olivier Favre

Same problem on Debian Jessie running KDE. Another similar Slack client does not have this issue though.

Yes, that's many changes, plus it seems that a few core classes have been removed along with the delete-by-query feature. I think rebasing update-by-query on the delete-by-query plugin architecture may...

A new Reindex API has recently been merged inside elasticsearch. It's still inside a feature branch for now, though. https://github.com/elastic/elasticsearch/pull/15125 It may help speeding up the rewrite a bit by...

Have you tried tweaking the `action.updatebyquery.bulk_size` elasticsearch configuration option?

Do you still face the same problem with v2.2.0 of the plugin and ES 1.3?

Why would setting the `bulk_size` to the bulk `queue_size` help? The former is about the number of documents within a single bulk request, whereas the latter is about the number...

The plugin works as follows: - Upon reception of a new request, the involved shards are listed and the per-shard request is either being processed locally or remotely. All this...

I've integrated your remarks in the [readme](https://github.com/yakaz/elasticsearch-action-updatebyquery#configuration-options), thanks! Your advice to lower the `bulk_size` has been left-out, as I think this is a bad idea performance-wise. I hope we'll sort...

Lowering the `bulk_size` may help in some scenarii. But I still see no direct relation between the value of this plugin's `bulk_size` and Elasticsearch's bulk `queue_size`. Explanation: Adding one long...

@consulthys Could you test with a lower value for `bulk_size`? Like I previously suggested, `1` and "total hits" are two extreme values that may help to better grab the inner...