Titan Combat 1------Introduction

send an article commemorating Kobe Bryant's retirement.

I. Use version

So far, the Titan mainstream version has 2. One is titan1.0.0 and the other is titan0.5.x. titan1.0.0 is currently the most active version. titan0.5.x is no longer updated, and the latest version is titan0.5.4. : Https://github.com/thinkaurelius/titan/wiki/Downloads. As you can see from the download page, titan1.0.0 combines TINKERPOP3.0.1,TINKERPOP3 and 2 with a big difference, and 3 is to combine all the components in 2 into one gremlin to use. The use of the more convenient, but also has a lot of new features. However, compared to the regret is that titan1.0.0 does not support hadoop2.x well, its testing and use of the version are hadoop1.x, but most enterprises are using the hadoop2.x version. Such titan1.0.0 cannot be used, so if the enterprise is using hadoop2.x, only the titan0.5.x version can be used so far.

Here are two images to compare the similarities and differences between the titan0.5.4 and titan1.0.0 versions:

The above figure is titan0.5.4 version, the following figure is the titan1.0.0 version, can be very intuitive to see the differences. Version 0.5 is used in conjunction with the tinkerpop2.x version, while the 1.0 version is used in conjunction with the tinkerpop3.x version.

Tinkerpop2.x:http://www.tinkerpop.com/docs/wikidocs/index.html

tinkerpop3.x:http://tinkerpop.incubator.apache.org/docs/3.0.1-incubating/

In addition, the titan0.5.4 version does not extend to the calculation and analysis (OLAP) of graphs, but only the OLTP phase of the data. And titan1.0.0 has a great breakthrough in the calculation and analysis of graphs.

Note: The Titan version used is 0.5.4.

Two. Bottom-level storage and Index backend

Titan's underlying storage has three optional databases: Cassandra, HBase, BerkeleyDB, where Cassandra and hbase are distributed databases, and BerkeleyDB does not support distribution.

Note: The underlying storage is used for the HBase database.

Titan has three selectable index backend, namely: Elasticsearch, SOLR, Lucene,elasticsearch and SOLR are all based on Lucene's upper level project.

Note: The index backend used is elasticsearch.

Titan Combat 1------Introduction