cassandra-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sylvain Lebresne <sylv...@datastax.com>
Subject Re: March 2015 QA retrospective
Date Fri, 10 Apr 2015 09:19:01 GMT
On Thu, Apr 9, 2015 at 10:05 PM, Tyler Hobbs <tyler@datastax.com> wrote:

>
> In general, though, the dtests don't have great coverage on prepared
> statements.  We could add an abstraction layer over query execution to use
> prepared vs unprepared statements with an environment variable.
>

I'll note that the CQL unit tests do that (being able to switch between
prepared vs unprepared with an environment variable) so what we should
really
do is migrate the dtests to unit tests (which has been discussed already).
I've created CASSANDRA-9160 to keep track of this (I though we had a ticket
for
that already but haven't been able to find it).


>
> Of course, we could also ensure all
> (applicable) tests cover both COMPACT and non-COMPACT, but other table
> settings like the compaction strategy and caching could also affect some
> tests.
>

Similarly to what's above, it should be too hard to have the CQL unit tests
automatically use COMPACT tables when some environment variable switch is
providen. We could also implement some kind of randomization of parameters
so
that tests don't always run with the default settings. I've created
CASSANDRA-9159 to track that.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message