Huawei Hosts First Apache HBaseCon Asia

In early August, this year, Huawei held the first Apache HBaseCon Asia in Shenzhen, China.

The conference brought together over 300 HBase developers from well-known companies in the industry. A number of heavyweight Apache HBase Project Management Committee (PMC) members and committers, as well as more than 20 HBase senior engineers from China, the United States, India, and other countries delivered a total of 19 excellent speeches.

They shared their experiences in HBase 2.0, key kernel features, HBase application cases and industry innovations, cloud-based HBase services, and container technology.

The conference focused on HBase 2.0 planning and key features. HBase 2.0 is expected to be released at the end of 2017. HBase 2.0 has resolved 4400 issues, and its key features and improvements include a New Region Assignment Manager, off-heap read/write path, in-memory compaction, a RegionServer group, and a new Async Client.

The upcoming HBase 2.0 delivers significant improvements over HBase 1.x. These improvements cover stability, throughput, concurrency, and Mean Time to Repair (MTTR).

Four speakers from Huawei discussed application innovations in HBase-related sectors and shared their latest ideas on multi-tenancy, backup, and disaster recovery. Their speeches on the HBase application innovations focused on two aspects. One is CTBase, a light-weight HBase client for structured data which can simplify HBase operations for structured data storage.

The other aspect is Tagram, a distributed bitmap index implementation for HBase. It is mainly used to store persona attribute data, and offers millisecond-level latency for attribute-based ad-hoc queries.