Releases: flipkart-incubator/hbase-orm
Releases · flipkart-incubator/hbase-orm
v1.19
- [Enhancement] Support for asynchronous access to HBase (via AsyncConnection, AsyncTable, AsyncAdmin etc.) -- You get a
CompletableFuture
of result of your action, instead of blocking the thread. - [Change] Upgrade to HBase v2.2
v1.18
- [Enhancement] First-class support for namespaces.
HBTable
annotation now acceptsnamespace
as a parameter (this is optional, to maintain backward compatibility with older versions) - [Enhancement] Upgraded dependent libraries and maven plugins
- [Enhancement] Added Jackson datatype support for JDK 8 types by default
- [Fix] Fixed a rare issue which occurred if you were instantiating object mapper with more than one codec
- [Improvement] Error handling, Javadoc and README improvements
- [Change] Moved HBase library dependency from Cloudera's Hortonworks Data Platform version to Apache version
v1.16
- [Enhancement] Added
exists
methods to check whether a row exists or not (by single row key and by array of row keys). - [Change]
persist
anddelete
methods now takeT
whichextends HBRecord<R>
, instead ofHBRecord<R>
directly (change is backward compatible). - [Change] Removed obsoleted annotation
@HBRowKey
. - [Improvement] Maven test dependencies: Upgraded JUnit and Mockito. Removed
hadoop-mapreduce-client-core
as dependency. Upgradedmaven-javadoc-plugin
. - [Change]
fetchFieldValues(R[] rowKeys ...)
methods now return values of field in the same order as that of array (change is backward compatible). - [Improvement] Improved javadoc. Improved documentation.
v1.15
- [Improvement] Slightly improved performance of 'gets' and 'puts' by avoiding redundant class validations.
- [Improvement] In case of large row keys, slightly improved performance by avoiding a deep copy.
- [Library Upgrade] Jackson version upgraded to 2.9.10.4.
- [Fix] In the remote case of accessing an HBase table via two classes (one that uses
HBColumn
and one that usesHBColumnMultiVersion
for the same field), pick up the latest value instead of earliest.
v1.14.1
v1.14
- [Enhancement] New
HBAdmin
class for DDL operations (create table, delete table etc.). - [Improvement] Rewrote
BestSuitCodec
(default codec of this library) to remove reflection. Performance gain of at least 20% in serialization/deserialization. - [Change] Deprecated
HBRowKey
annotation. This will be removed in upcoming versions. - [Improvement] Migrated test cases to JUnit 5.
v1.13
- [Enhancement] Introduced
byPrefix
methods for fetching records by row key prefix - [Enhancement] Introduced
.records
methods for fetching large number of records (these methods efficiently maintain iterators internally without loading records in memory - hence safe to fetch millions of records using them) - [Enhancement] Introduced method to fetch records by HBase's
Scan
object - [Improvement] Improved build time by optimizing test cases
- [Improvement] Further improved test cases, javadoc and documentation
v1.12.1
v1.12
- [Enhancement] Functionality to append column values, enabled through a bunch of
append
methods on DAO class. - [Enhancement] DAO class is now fully thread-safe. You can now have just one DAO class per entity for the entire lifecycle of your application.
- [Enhancement] Upgraded to HDP 3.1. Replaced internal usage of all obsolete methods with newer (better) alternatives.
- [Improvement] Version updates of underlying libraries (e.g. Jackson) and maven plugins
- [Fix] Closed
ResultScanner
stream (in two bulk methods, they were left open) - [Improvement] Improved javadoc and error messages
v1.11-hdp31-78
- [Enhancement] Upgrade to HDP v3.1