[
https://issues.apache.org/jira/browse/GORA-497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16110446#comment-16110446
]
ASF GitHub Bot commented on GORA-497:
-------------------------------------
Github user renato2099 commented on the issue:
https://github.com/apache/gora/pull/110
Hi @madhawa-gunasekara ,
Sorry for the late reply, but to handle that case why don't just do the
mapping manually as we currently do and only serialize differently? So what I
had in mind was using our own same old avro java bean for keeping things
in-memory, but when pushing this into Cassandra, then (1) store everything as
avro binary data (serialize every type as binary data as we did), or (2) use
cassandra native serialization so users can use regular CQL for checking their
data.
You accomplished the same :) BUT now we have also introduced a different
in-memory data bean which will not help in keeping a single entry-point for all
no-sql databases (now we need Cassandra java beans). I think we could just go
back to use our own avro data beans and use the native serialization you have
coded, and then later handling the avro binary serialization. What do you
think? how much work would it be?
> Migrate CassandraThrift to CQL
> -------------------------------
>
> Key: GORA-497
> URL: https://issues.apache.org/jira/browse/GORA-497
> Project: Apache Gora
> Issue Type: Improvement
> Components: gora-cassandra
> Reporter: Madhawa Gunasekara
> Assignee: Madhawa Gunasekara
> Labels: gsoc2017
> Fix For: 0.8
>
>
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
