Pivotal Greenplum database products is based on MPP architecture, it can meet the demand for large-scale analysis task the next generation of data warehouse, by automatically the data partition and parallel query execution multi-node, Greenplum make a database that contains hundreds of node cluster to run as stand-alone version of the traditional database is simple and reliable, at the same time provides several times or even one hundred times performance improvement, in addition to the traditional SQL, Greenplum also supports graphs, the text index, stored procedure many analysis tools, such as
Greenplum 5.1.0 can be downloaded from here (https://network.pivotal.io/products), the document here (https://gpdb.docs.pivotal.io/510/main/index.html), the home page here (http://greenplum.org/), the source code in making (https://github.com/greenplum-db/gpdb),
new features to support
enhanced GPORCA short on the performance of the query optimization
To do not need to estimate the field during the Greenplum 5.1.0 skip statistics data query and generate, reduce the optimization time consuming, for short query performance significantly, in previous versions, even if only the width of the required fields (width) information, GPORCA will query field of statistics and other
Promoted GPORCA optimizer performance
Greenplum 5.1.0 adds the following GPORCA performance enhancements
1. To a large number of tables associated operation generates the execution plan, GPORCA reduced in estimate the number of the greatest join combination, has little effect on the improved query performance, it greatly reduces the optimization time, in previous versions, GPORCA will for every possible combination of the join estimate to determine the optimal solution, thus take longer to generate execution plan
2. To include associations sub-query (correlated subquery) and the subquery contains the query window function, their GPORCA be generated based on the join more efficient query plan,
GPORCA can support the partition table index of child node
In Greenplum 5.1.0, if a partition table data node contains index, GPORCA child nodes in the generated for this data query plan, will use the index information; The previous version will not use the index of leaf nodes,
support between the table and external program COPY
Greenplum 5.1.0 support the copy in the Postgres 9.3 to/from the program function, the user can specify the external commands, run in parallel on each segment, handle the copy command output or input data to the copy command,
gptransfer supports the SHA - 256 data validation
Greenplum 5.1.0, gptransfer validated SHA265 when transmitting data according to the data, when the operating system you enable FIPS md5 checksum is considered unsafe algorithm, therefore gptransfer using more advanced SHA265 calibration algorithm,
The performance of the promoted gprecoverseg
Greenplum 5.1.0, when segment node has a number of files, to gprecoverseg operation, performance has improved,
new external data engine PXF
In introducing new external Greenplum 5.1.0 data frame PXF (Pivotal extention framework), it is deployed on each Segment of the physical machine operation, provides to the HDFS file system with the support of the HIVE, PXF external data provides an abstract interface, easy to support a wide variety of data sources,
test feature
Besides the function of the formal support, Greenplum 5.1.0 also includes the following several test features:
Recursive CTE
CTE (Common Table Expressin) defines a repeat in the same query can be used by the temporary Table, can greatly simplify the SQL statement, Greenplum 5.1.0 definition support the recursive CTE in keywords, allowing the CTE definition can be recursive references,
Resource management based on the Resource group
The Resource group is the next generation of Greenplum resources management framework, which can be used to manage the number of simultaneous queries, and each query allows the use of the limitation of CPU and memory, Greenplum 5.1.0 default still use the old version of the Resource management mechanism, can modify gp_resource_manager as "group" to try out new Resource group function,
Pgadmin4 support
Greenplum 5.1.0 compatible PGadmin4, the user can query to browse through PGadmin4 Greenplum list (including AO table) and DDL information,
Greenplum 5.1.0 extension component
Client end tool
extension module
other extensions
PXF Extension Framework
In introducing new external Greenplum 5.1.0 data frame PXF (Pivotal extention framework), it is deployed on each Segment of the physical machine operation, provides to the HDFS file system with the support of the HIVE, PXF external data provides an abstract interface, easy to support a wide variety of data sources,
Greenplum - Spark Connector
Support Greenplum with high speed parallel data transmission between the Spark,
Pivotal GPText
Pivotal Greenplum Database 5.1.0 can support GPText version 2.1.3 and later version, is Greenplum GPText provides text search engine, can support full-text retrieval and text analysis function,
Greenplum 5.1.0 support platform
Greenplum server supports the following platform
Red Hat Enterprise Linux 64 - bit 7 x
Red Hat Enterprise Linux 64 - bit 6. X
SuSE Linux Enterprise Server 64 - bit 11 SP4
CentOS 64 - bit 7 x
CentOS 64 - bit 6. X
Greenplum Java components depend on the following version of Java
8. XXX
7.
XXX
Greenplum dependence on runtime requires the following packages
OpenSSL 1.0.2 l (with FIPS 2.0.16)
7.54 cURL the OpenLDAP 2.4.44
Python 2.7.12
Client end tools can support the following platform
Red Hat Enterprise Linux 64 - bit 7 x
Red Hat Enterprise Linux 64 - bit 6. X
SuSE Linux Enterprise Server 64 - bit 11 SP4
CentOS 64 - bit 7 x
CentOS 64 - bit 6. X
Windows
AIX