Logstash performance degradation after upgrade

we were getting 60-80K indexing rate in our production cluster, after upgrading the cluster (from v 2.2.2) we had this go down to 12-14K.

Elasticsearch 2.3.1
Logstash 2.3.1
Marvel 2.3.1
Shield 2.3.1
Kibana 4.5.0

There is no change in any of config files. Is there a change in the logstash uses resources from 2.2.2 to 2.3.1

No information in the logs even in --debug mode

Any help in this direction is appreciated.

Looking at the 2.3 release notes this bullet stands out:

Reverted default output workers to 1. Previously we had made output workers the same as number of pipeline workers (#4877).

yes..That was the issue