Apache HBase Agent for Gluesync: Features Overview

Index

Source data from Apache HBase

Target Apache HBase

Agent changelog

Core principles

Communication between this agent and Apache HBase is achieved through its Java SDK.

Change data capture

Gluesync uses the scan feature provided by Apache HBase to catch changes occurring at the database level.

Platform supported features

Feature Support Notes

Feature	Support	Notes
Target role	✅	Writes to HBase using the native Java client.
Source role	✅	Uses HBase `scan` operations to read source tables for CDC.
Snapshots	✅	Full-table scans replicate snapshot data before CDC.
CDC	✅	Incremental scans detect new/updated rows between offsets.
CDC type	Scan-based polling	Relies on lightweight range scans with row-key bookmarking.
Bulk load	N/A	HBase client already batches writes; no staging-table workflow.
Source persisted internal cache	N/A	Offsets/row keys stored in Core Hub metadata only.
Target persisted internal cache	N/A	Writes go directly to HBase via the Java SDK.
Driver / connector	HBase Java SDK	Gluesync bundles the official client libraries.
Chains	N/A	Not supported being it a NoSQL database.
Truncate operation support (reading)	N/A	Scan-based CDC does not emit TRUNCATE events.
Truncate operation support (writing)	N/A	Gluesync does not issue table truncations automatically.

Target role

✅

Writes to HBase using the native Java client.

Source role

✅

Uses HBase scan operations to read source tables for CDC.

Snapshots

✅

Full-table scans replicate snapshot data before CDC.

CDC

✅

Incremental scans detect new/updated rows between offsets.

CDC type

Scan-based polling

Relies on lightweight range scans with row-key bookmarking.

Bulk load

N/A

HBase client already batches writes; no staging-table workflow.

Source persisted internal cache

N/A

Offsets/row keys stored in Core Hub metadata only.

Target persisted internal cache

N/A

Writes go directly to HBase via the Java SDK.

Driver / connector

HBase Java SDK

Gluesync bundles the official client libraries.

Chains

N/A

Not supported being it a NoSQL database.

Truncate operation support (reading)

N/A

Scan-based CDC does not emit TRUNCATE events.

Truncate operation support (writing)

N/A

Gluesync does not issue table truncations automatically.

Supported version

This agent has been tested under HBase 2.4.15.

Q&A

I have got Phoenix running on top of my HBase deployment, is it supported? Sure, Gluesync doesn’t make use of any other component other than plain HBase. You can still make use of Phoenix while Gluesync migrates and offloads your data from the datastore.

My company has more than a Petabyte of data inside our deployment, does it work with Gluesync? How many resources do I have to dispose of to ensure it will work well? Resource amount is based on a few key driving factors:

Network bandwidth;
Storage performance;
CPU;
RAM.

We will take care of gathering these KPIs and provide you with the proper sizing of the needed resource allocation for the Gluesync instances. There is no amount of data that scares Gluesync: when properly sized your data will smoothly flow into the other end, consistently and safely.