# Updating or modifying parameters in river plugin **URL:** \ **Category:** Elasticsearch\ **Created:** [June 25, 2012, 1:42pm UTC](https://discuss.elastic.co/t/updating-or-modifying-parameters-in-river-plugin/8213 "2012-06-25T13:42:56Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![sujoysett](https://avatars.discourse-cdn.com/v4/letter/s/2acd7d/32.png) [@sujoysett](https://discuss.elastic.co/u/sujoysett)\ **Post date:** [June 25, 2012, 1:42pm UTC](https://discuss.elastic.co/t/updating-or-modifying-parameters-in-river-plugin/8213/1 "2012-06-25T13:42:56Z")
Hi, Facing one problem while modifying tracks or users, streaming parameters gets changed only on node restart. The data in the index "\_river" is getting changed, but data in actual stream doesn't change until node restart. Kimchy had suggested deletion and re-creation of the index "\_river" for updates to reflect in streaming ( [https://github.com/elasticsearch/elasticsearch-river-twitter/issues/9](https://github.com/elasticsearch/elasticsearch-river-twitter/issues/9)). Can't we have a rest kind of endpoint as an addition over the \_river (say for example, \_river/{index\_name}/\_reconnect ), which will explicitly do the task performed by the reconnect() method in the code? Or simpler still, why doesn't a re-connection get queued on changing the \_river/{index\_name}/\_meta endpoint once again after the initial? Since there are limits on the tracks or users in the public twitter stream, it becomes specially necessary to delete unnecessary information, and adding new tracks/ users to existing is always a necessity. Thanks in advance, Sujoy. ---
**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\ **Post date:** [July 6, 2017, 3:22am UTC](https://discuss.elastic.co/t/updating-or-modifying-parameters-in-river-plugin/8213/2 "2017-07-06T03:22:40Z")
Closed on July 6, 2017, 3:22am UTC