NebulaGraph RC4 Release Note


This release introduces some new features such as INDEX where you can use the LOOKUP ON syntax to query indexed data, access management based on SPACE, and Nebula Stats Exporter to interact with Grafana and Prometheus.

By the end of this release note you will find the steps to upgrade from RC3 to RC4.

Below is a list of features and tools in detail:

New features

  • Support Index to make searches of related data more efficient. An index on a property/properties combination can be created with CREATE INDEX. DROP INDEX is to drop an index #1776.  REBUILD INDEX is to reindex the data #1566.
  • Support LOOKUP ON  to query on index #1705. See #1738 for the performance of Storage Engine when inserting data with an index.
  • Account management and access control #1842#1873 
  • All users may perform only the operations permitted to them. For roles available in NebulaGraph and their privileges please refer to  #1929. Add --enable_authorize=true to nebula-graphd.conf and restart the services to enable authentication.
  • Support TTL to remove items after a certain amount of time automatically #1584#422, #1934
  • Enhance DELETE VERTEX  to support deleting a batch of vertices. And supports hash() and uuid()  functions for VertexID #1317#1759
  • Introduce Job Manager to manage the jobs that take a long time of Storage Engine. At present, it supports flush  and compact operations. SUBMIT JOB  is to submit a job.  STOP JOB stops the running jobs. SHOW JOB shows the detailed info of a job. RECOVER JOB is to put back the failed job to the queue. #1424
  • Support BIDIRECT for GO query to traverse along with both incoming and outgoing directions.  #1740#1752
  • Support Reservoir Sampling. Set enable_reservoir_sampling to TRUE is to turn on the sampling. Max_edge_returned_per_vertex in nebula-storage.conf is to configure the number of returned edges. #1746#1915
  • Support more character sets and collations.  SHOW CHARSET and SHOW COLLATION statements show all available character sets and collations. It can be configured when creating the space. The default CHARSET is utf8, and corresponding COLLATE is utf8_bin. #1709

OLAP Interface

  • Add Spark example for NebulaGraph. #56



  • RC4 is not compatible with the Data inserted before commitID 43453a0 (2020.02.06) due to the change of the underlying data structure.
  • The documentation has been moved to repo vesoft-inc/nebula-doc for better maintenance
  • For Nebula Python Client, is_async is not supported when creating ConnectionPool. Async client will be available later.

Upgrade from RC3 to RC4

If you are currently using a previous version of NebulaGraph and would like to try the new features in RC4, please follow the steps below for an upgrade:

  1. Stop all Nebula services
    1. Execute scripts/nebula.service stop all on each machine
    2. Execute scripts/nebula.service status all to confirm all services are stopped
  2. Install the new RPM package on each machine according to the OS you are using
    1. Get package: https://github.com/vesoft-inc/nebula/releases/tag/v1.0.0-rc4
    2. Install package, like: rpm -Uvh nebula-1.0.0-rc4.el7-5.x86_64.rpm
  3. Start Nebula
    1. Execute scripts/nebula.service start all on each machine
    2. Execute scripts/nebula.service status all to confirm Nebula services are started on each machine
  4. Reload your data into NebulaGraph

If you have any thoughts or suggestions, please feel free to leave us a comment via the Discussion board below! Or you may raise an issue directly on GitHub.

You might also like