It sounds good to do. One of the reasons to do sentry 2.0 was the integration with hive-authz2, but this is taking time for quality concerns. Perhaps is good to cut the branch for 1.9 now before starting to add other incompatible changes to the master branch.
+1 to release 1.9 + ha work. On Wed, Oct 4, 2017 at 5:19 PM, Kalyan Kumar Kalvagadda < kkal...@cloudera.com> wrote: > Hi all, > > I understand that the major areas that still needs significant amount work > on sentry 2.0 are > > 1. Integrating with Hive 2.0 > 2. Updating the java version > 3. Integrating with Solr-7 > > > *Integrating with Hive 2.0: *While testing stuff with Hive locally, we > found couple of issues with Hive-2 and sentry 2.0 integration using auth-v2 > interface. There were couple of fixes that are needed from Hive side, some > of them are already fixed in Hive 3.0 and some are yet to be fixed. Bottom > line is that we need some help from Hive community in fixing some issues > and back-porting them to some Hive-2.x release before we actually try to > continue testing it with new Hive-2.x release. This could take quite some > time > > *Integrating with Solr-7: *We happened to talk to contributors of this > feature. They might some time to actually get it ready. > > These outstanding work need not delay the release of sentry-HA which is > well tested locally at cloudera. This feature is something that the > community is waiting for quite some time. Why don't we release current > master as sentry 1.9 release? > > Sentry 1.9.0 if released, would below features. > > 1. Sentry-HA > 2. Integration with Hive 2.0 using auth-V1. > > Seeking inputs from every one in this regard. > > > -Kalyan >
