I strongly support this. I've seen many users and customers accidentally lose data because they didn't know this configuration exists. I always support defaulting to "Not lose data" and let the users who prefer high availability do some extra research. This will need to be very well documented - both the default trade-off, the alternative and the fact that things changed.
On Tue, Jan 3, 2017 at 10:54 AM, Ben Stopford <b...@confluent.io> wrote: > Hi All > > Please find the below KIP which proposes changing the setting > unclean.leader.election.enabled from true to false. The motivation for this > change is that it catches out new Kafka users who don’t realise the default > favours availability over data loss. > > This would mean clusters wishing to continue with unclean leader election > enabled would need to add the appropriate configuration on upgrade. > > Please let me know if you foresee any issue with this change, agree or don’t > agree. > > https://cwiki.apache.org/confluence/display/KAFKA/%5BWIP%5D+KIP-106+-+Change+Default+unclean.leader.election.enabled+from+True+to+False > > <https://cwiki.apache.org/confluence/display/KAFKA/[WIP]+KIP-106+-+Change+Default+unclean.leader.election.enabled+from+True+to+False> > > Thanks > > B > > Ben Stopford > Confluent, http://www.confluent.io <http://www.confluent.io/> > > > -- Gwen Shapira Product Manager | Confluent 650.450.2760 | @gwenshap Follow us: Twitter | blog
