[
https://issues.apache.org/jira/browse/BEAM-301?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15684238#comment-15684238
]
Tyler Akidau commented on BEAM-301:
-----------------------------------
You may already be aware of these, Neelesh, but I'd like to reference some docs
here that are relevant to this effort here in case you aren't (or for others).
I've been chatting on and off w/ the Calcite folks regarding SQL this year, in
particular regarding adding robust streaming semantics into the Calcite
algebra. We don't need any additions to get something basic, similar to what
Apache Flink has (and there's plenty of work around just defining how schemas
are mapped, how SQL is expressed in the pipeline, etc). But we will want them
some day.
At any rate, the relevant docs so far that I'm aware of:
1. Streaming SQL - Julian Hyde’s initial manifesto on integrating streaming in
Calcite SQL: https://calcite.apache.org/docs/stream.html
2. Triggers in Streaming SQL - My first attempt at tackling out-of-order
processing semantics for streaming SQL:
https://docs.google.com/document/d/1tSey4CeTrbb4VjWvtSA78OcU6BERXXDZ3t0HzSLij9Q/edit
3. Streams, joins and temporal tables - Julian’s proposal for tables which are
parameterized on a timestamp value:
https://docs.google.com/document/d/1RvnLEEQK92axdAaZ9XIU5szpkbGqFMBtzYiIY4dHe0Q/edit
4. Relational Queries on Data Streams - Apache Flink proposal for standard SQL
based queries on data streams:
https://docs.google.com/document/d/1qVVt_16kdaZQ8RTfA_f4konQPW4tnl8THw6rzGUdaqU/edit
I've been wanting to write up one more doc trying to reconcile #2 with #3, but
haven't gotten around to finishing it yet.
> Add a Beam SQL DSL
> ------------------
>
> Key: BEAM-301
> URL: https://issues.apache.org/jira/browse/BEAM-301
> Project: Beam
> Issue Type: New Feature
> Components: sdk-ideas
> Reporter: Jean-Baptiste Onofré
> Assignee: Neelesh Srinivas Salian
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
