You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by "Birdsall, Dave" <da...@hp.com> on 2015/05/04 18:12:24 UTC
[ANNOUNCE] Announcing Trafodion 1.1
Good morning HBasers,
I'm pleased to announce that Trafodion Release 1.1 is now available for download at http://downloads.trafodion.org/. This is a minor release, improving upon Trafodion's position of being the Enterprise-class Transactional SQL Engine for HBase and Hadoop. Even though we've classified it as a minor release, we were still very busy.
Here is a summary of the production-ready features in this release:
1. Support for HDP 2.2 and CDH 5.3 distributions
2. Several performance features, e.g., improved plan quality, query caching to reduce compilation time, and Skew Buster, which is patented technology to create better query execution in the presence of skewed data values
3. Improvements in high availability
4. An offline backup/restore facility
5. A parallel data loader/extractor tool for trickle-loading to and extraction from Trafodion tables, including the ability to extract from many popular databases and insert into Trafodion in parallel
6. Support for the HP Data Services Manager (HP DSM), a tool that provides unified, browser-based management of Hadoop, Trafodion and HP Vertica data services
7. Improved repository support
8. Extensions to query cancel, to support cancelling more kinds of statements
9. Improved security support
10. Improved installer
We also have several "Technology Preview" features in this release. These are features that are (sometimes partially) code complete but not yet fully tested. You're welcome to try them out and tell us what you think:
1. The ability to create table statistics via sampling during table load (faster than loading a table and doing UPDATE STATISTICS afterward)
2. Several improvements to transaction processing, including local transactions (an efficiency improvement), Stateless/Stateful Concurrency Control (SSCC) (an improvement in isolation over traditional MVCC), and transactional support for DDL operations
3. A metadata cleanup utility (useful when a non-transactional DDL operation fails)
4. Table-mapping user-defined functions (TMUDFs), that is, support for user-defined functions that can take a table as input and produce a table as output
We fixed 96 defects in this release as well.
You can read more here: https://wiki.trafodion.org/wiki/index.php/Release_Notes_(1.1.n).
Happy reading,
Dave
PS: We'd LOVE to have more contributors! We'll soon be proposing Trafodion as an incubation project to the Apache Software Foundation, so we'll be using tools you're likely familiar with. Inquire at our website, http://www.trafodion.org, or by e-mail at trafodion@hp.com<ma...@hp.com>.
PPS: Connect with us and see our talks at HBaseCon in San Francisco on May 7<https://wiki.trafodion.org/wiki/index.php/Event:2015/05/07_HBaseCon_2015>.
Re: [ANNOUNCE] Announcing Trafodion 1.1
Posted by Stack <st...@duboce.net>.
Congrats Trafodion. Thanks for posting notice here.
St.Ack
On Mon, May 4, 2015 at 9:12 AM, Birdsall, Dave <da...@hp.com> wrote:
> Good morning HBasers,
>
> I'm pleased to announce that Trafodion Release 1.1 is now available for
> download at http://downloads.trafodion.org/. This is a minor release,
> improving upon Trafodion's position of being the Enterprise-class
> Transactional SQL Engine for HBase and Hadoop. Even though we've classified
> it as a minor release, we were still very busy.
>
> Here is a summary of the production-ready features in this release:
>
>
> 1. Support for HDP 2.2 and CDH 5.3 distributions
>
> 2. Several performance features, e.g., improved plan quality, query
> caching to reduce compilation time, and Skew Buster, which is patented
> technology to create better query execution in the presence of skewed data
> values
>
> 3. Improvements in high availability
>
> 4. An offline backup/restore facility
>
> 5. A parallel data loader/extractor tool for trickle-loading to and
> extraction from Trafodion tables, including the ability to extract from
> many popular databases and insert into Trafodion in parallel
>
> 6. Support for the HP Data Services Manager (HP DSM), a tool that
> provides unified, browser-based management of Hadoop, Trafodion and HP
> Vertica data services
>
> 7. Improved repository support
>
> 8. Extensions to query cancel, to support cancelling more kinds of
> statements
>
> 9. Improved security support
>
> 10. Improved installer
>
> We also have several "Technology Preview" features in this release. These
> are features that are (sometimes partially) code complete but not yet fully
> tested. You're welcome to try them out and tell us what you think:
>
>
> 1. The ability to create table statistics via sampling during table
> load (faster than loading a table and doing UPDATE STATISTICS afterward)
>
> 2. Several improvements to transaction processing, including local
> transactions (an efficiency improvement), Stateless/Stateful Concurrency
> Control (SSCC) (an improvement in isolation over traditional MVCC), and
> transactional support for DDL operations
>
> 3. A metadata cleanup utility (useful when a non-transactional DDL
> operation fails)
>
> 4. Table-mapping user-defined functions (TMUDFs), that is, support
> for user-defined functions that can take a table as input and produce a
> table as output
>
> We fixed 96 defects in this release as well.
>
> You can read more here:
> https://wiki.trafodion.org/wiki/index.php/Release_Notes_(1.1.n).
>
> Happy reading,
>
> Dave
>
> PS: We'd LOVE to have more contributors! We'll soon be proposing Trafodion
> as an incubation project to the Apache Software Foundation, so we'll be
> using tools you're likely familiar with. Inquire at our website,
> http://www.trafodion.org, or by e-mail at trafodion@hp.com<mailto:
> trafodion@hp.com>.
>
> PPS: Connect with us and see our talks at HBaseCon in San Francisco on May
> 7<https://wiki.trafodion.org/wiki/index.php/Event:2015/05/07_HBaseCon_2015
> >.
>
>
>