HBase to Cassandra Migrations
HBase can be positioned as almost an original part of the Hadoop stack, which is a great competitive advantage, but when viewed with closer scrutiny, it becomes very clear that HBase is built for analytics and not online environments – good for online analytics (OLAP) but not online transactions (OLTP) as well as analytics. This is becoming a core requirement of todays business applications. Add to this the fact that HBase is extremely complex operationally as well as a lack of support for multiple datacenters and the challenges that must be overcome start to add up. This is why Cassandra has been the choice for companies that need to rely on more from their database.
Use Cases and Migration Resources
Wattgo needed a low latency solution and HBase could not meet their performance requirements for reads and writes; migrating from HBase to Cassandra better handled their time series data.
AppScale’s preference of Casssandra over HBase comes down to Cassandra’s ability to handle high write volume as well as ease of configurability.
Cloze offers a ‘noice-cancelling’ inbox to hundreds of thousands of users and relies on Cassandra as an always available, secure, high volume data store.
Digital Reasoning leans on Apache Cassandra for high volume data ingestion and multi-datacenter support
eBuddy’s messaging platform went with Cassandra as their primary database platform because it’s written in Java, it’s open source, it’s easy to add and remove nodes, and the ease of configuration was immensely useful.
eComNext decided on Apache Cassandra instead of HBase as their datastore of time series data thanks to the faster and simpler setup as well as low cost of maintenance.
General Sentiment’s marketing intelligence solution depends on Cassandra to host their text analytics based on ease of setup and high performance.
GX Software uses Cassandra exclusively as their data store thanks to a simple deployment model, storage capable of an extremely high volume of simultaneous reads/writes (1 billion + user records), linear scalability and no single point of failure.
Health Market Science, an integrated service provider to the health care industry, decided to deploy Cassandra thanks in part to the stronger code base and highly active community.
Loop Logic’s media sharing platform went with Cassandra instead of HBase as their real-time analytics platform for its ease of setup as well as unmatched scalability.
Ok.ru (Odnoklassniki), the largest Russian speaking social network, is powered by Cassandra with 23 clusters and 370 nodes in production due to ease of use and maintenance.
Outbrain opted to deploy Apache Cassandra over HBase thanks to Cassandra’s multi datacenter replication and low operational overhead.
SkillPages employment neworking site runs on Apache Cassandra instead of HBase due to linear scalability, ease of deployment and multi datacenter support.
Tuato calls their decision to use Cassandra as their datastore for their secure mail application instead of HBase a ‘no brainer’ thanks to the ease of use, scalability and no single point of failure.
The slideshare Comparing Architecutres: Cassandra versus the Field compares Cassandra to other NoSQL stores, especially HBase so you can pick the right database for your project.
Apache Cassandra at GumGum — From HBase to Apache Cassandra from the LA: Video from the LA Cassandra Users Group:
GumGum relies on Cassandra heavily for storing image metadata. After using HBase for 4 years, they decided to make the switch. Why did they do that? What problems did they face? What was their migration journey? What were the mistakes they made? What were the choices they made when hosting Cassandra on EC2? This presentation will narrate GumGum’s exciting journey to Apache Cassandra.