calcite-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Julian Hyde <>
Subject Re: Sequence support
Date Fri, 11 Aug 2017 05:40:12 GMT
I don’t know what you mean by “operations”. Most relational operators can be pushed down.
And a few contributors are improving the JDBC adapter all the time.

Sequences are only partially done. Most of the work has been done by and for the Phoenix project,
which implements sequences as tables. They are implemented in Phoenix’s extensions to Calcite,
but Calcite only has parse+validate support. It would be nice if someone got them working
so that they could be pushed down to a JDBC data source. (Not all DBMSes support them. I know
Oracle does, but I don’t think MySQL or MSSQL do.)

You are correct. Transactions are not implemented. Again, we’d need to push them down to
the source.


> On Aug 10, 2017, at 10:27 AM, Christian Beikov <> wrote:
> Hello,
> I am new to Calcite but I already got quite far with integrating it into a test project.
Now I am a bit stuck and I am not sure what is happening.
> I would like to use Calcite for data federation, but first I am trying to simply put
it between the application and the actual JDBC connection. I read something on the website
that operations aren't pushed down to JDBC schemas yet? That would be unfortunate as that
is something I actually need. Can you confirm what the status of that is?
> While trying to insert data, I noticed that SQL sequences don't seem to be supported.
Is there a special undocumented syntax for retrieving sequence values? I tried "next value
for SEQUENCE" as something in SqlKind was suggesting that might be the appropriate syntax
but it didn't work.
> Finally, I also noticed that transactions don't seem to work properly. Apparently CalciteMetaImpl#commit
throws an UnsupportedOperationException? Am I missing something here or is that a current
> Thanks in advance!
> -- 
> Mit freundlichen Grüßen,
> ------------------------------------------------------------------------
> *Christian Beikov*

View raw message