Is that version going to be stable for production?
Well, 1M$ question )). There is a lot of discussion on this topic. In terms of tick-tock releases odd releases (e.g. 3.11) are bug fixes. But I guess latest 3.0.x (3.0.14 now) should more stable. Best regards, Vladimir Yudovin, Winguzone - Cloud Cassandra Hosting ---- On Mon, 26 Jun 2017 07:09:08 -0400 Ali Akhtar <ali.rac...@gmail.com> wrote ---- Is that version going to be stable for production? I'm looking for something that I can just install, add nodes when needed, but otherwise not have to worry or think about, even if it means downgrading to a lower version and rewriting some of the code involving UDTs. On Mon, Jun 26, 2017 at 3:51 PM, Vladimir Yudovin <vla...@winguzone.com> wrote: Latest comment in this JIRA is "I've committed to 3.11". 3.11 change log also contains "* Fix validation of non-frozen UDT cells (CASSANDRA-12916)" (merged from 3.10) So try version 3.11 Best regards, Vladimir Yudovin, Winguzone - Cloud Cassandra Hosting ---- On Thu, 22 Jun 2017 10:17:15 -0400 Ali Akhtar <ali.rac...@gmail.com> wrote ---- I'm running cassandra 3.9, but it doesn't seem stable. E.g, one of my nodes recently crashed with the message 'org.apache.cassandra.db.commitlog.CommitLogReadHandler$CommitLogReadException: Unexpected error deserializing mutation; saved to /tmp/mutation3976606415170694683dat. This may be caused by replaying a mutation against a table with the same name but incompatible schema. Exception follows: org.apache.cassandra.serializers.MarshalException: Not enough bytes to read 0th field board_id' It looks like this particular bug is fixed in 3.10: https://issues.apache.org/jira/browse/CASSANDRA-12916 Is there a stable version with support for frozen UDTs that I should use? If not, should I change my UDT code to use text, and revert to a 2.x version which is stable? I'm still in development, so it will be a pain, but I can revert to non frozen UDTs.
