flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From delding <...@git.apache.org>
Subject [GitHub] flink issue #2332: [FLINK-2055] Implement Streaming HBaseSink
Date Tue, 27 Sep 2016 05:20:54 GMT
Github user delding commented on the issue:

    Hi @ramkrish86 , I'm thinking replace batch() with mutateRow() because it provides atomic
ordered mutations for a single row, but it only supports Put and Delete which should be fine
since only Put and Delete are idempotent, this way we can implement Put and Delete without
using WriteAheadSink (in case of deterministic processing). What do you think? 
    Regarding Append and Delete, as HBase doesn't support distributed transaction across multiple
rows, we might need WriteAheadSink and figure out a way to roll back table to the last checkpoint
state. I'm thinking about this right now.
    So it might make sense to have two HBaseSinks, one for Put/Delet, the other for Append/Delete
and non-deterministic processing.

If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.

View raw message