it’s all been done: release 1.1.0 is ready to be downloaded and used before you go on holidays! Thank you to the 3 committers and the always growing contributors for this fruit of the keyboard.
Before to go in deep with this release an announcement: NuvolaBase
, the company behind OrientDB, is working to the new “OrientDB Enterprise
” product planned for September 2012. OrientDB is and will always remain Open Source
, but companies asked us some “enterprise” features we want to provide as commercial product based on top of OrientDB server. So starting from the next release (1.2) we will have 2 lines of products:
- OrientDB community and
- OrientDB enterprise
- New distributed server architecture with balance of workload, full replication and management of failures
- New MMAP manager optimized for 64bit OS
- Index: new support for binary keys and collections in composite keys
- Support for in-memory database replication
- Object Database added support for cascade deletion, attach/detach, ENUM and binary data (ORecordBytes)
- SQL: new DROP CLUSTER command, INSERT now supports clusters, new CREATE VERTEX and CREATE EDGE commands
- Support for Gephi visual tool
- 50 bugs fixed
This is the most important feature of this release and allow to use OrientDB with multiple server nodes. To reach this goal OrientDB uses the Hazelcast
library that is licensed with the same OrientDB license: Apache 2.
Since the “old” replication commands are not more working the only way to get working a replication among multiple servers is to run OrientDB in distributed mode with all the servers having the same databases before to start.
To start it use dserver.sh
(or dserver.bat under MS Windows systems) instead of the classic server.sh/bat
” script takes config/orientdb-dserver-config.xml as configuration instead of
Contribute to the OrientDB project (sorry but I can’t stop to remember this)
Please help OrientDB to be more famous by writing a short review in your Blog, Magazines and Mailing Lists. The magic formula is: More users = More test = More stable = More support (drivers, plugins, etc).
In order to use the HW/SW resource at the maximum, starting from this release OrientDB brings in the distribution 3 new libraries to uses some low-level feature provided by the OS:
If these libraries are not available in the classpath, OrientDB will continue to use the standard Java runtime. This releases has a new MMAP manager optimized for 64bit systems. The older one remains available. To activate it set the system variable: file.mmap.useOldManager=true at startup.
Far to provide a 100% compliant JPA interface, this release adds the support for cascade deletion, detach/attach of POJO, ENUM and the management of binary data by using the ORecordBytes objects directly by the POJO.
If you’re a Hibernate/JPA user please don’t try OrientDB Object Database interface, otherwise you will never go back to a OR-Mapping tool! 🙂 Some users is experiencing the usage of OrientDB through the Object Database interface + JRebel in Web Applications: this is the maximum of productivity because they can change the POJO classes at run-time without stop the JVM and OrientDB gets all the new POJO classes in a flash. Cool!
To report an issue please follow the suggestions posted some days ago
, in particular attach the following information:
- OrientDB version. Please don’t forget this! If it’s a SNAPSHOT please tell if it’s from SVN (revision number) or from MAVEN
- Operative System
- 32-bit or 64-bit
- JVM version
- Dump of configuration and profiler
We’ve built a new Partnership Agreement
for companies interested to sell NuvolaBase or OrientDB services as “reseller” or “services partner”. Please contact firstname.lastname@example.org
to have more information about it.