SANTA CLARA, Calif. -- As Huawei lifts the curtain on its big data and open source strategy, it’s clear that Huawei makes strategic investment on Spark and its ecosystem and takes aim at the telecom industry and other verticals. Today hundreds (physical and online) attendees from Spark community gathered in Bay Area Spark Meetup at Huawei USA R&D Center. In Spark tech talks, Huawei highlighted one of community projects “Spark SQL on HBase”, designed to accelerate online data query and analytics for large data sets, which Huawei global team developed since late 2014 and open sourced recently, along with other new features in the machine learning area of Spark MLLib. Such move is in sharp contrast to traditional impression on the firm. “Huawei’s contributions to the Spark project and community prove its long term commitment and strong belief in Spark. As an early corporate sponsor of the UC Berkeley AMPLab, where we built Spark, Huawei has been supporting the project in a variety of ways for years.” Andy Konwinski, Spark committer and co-founder of Databricks, said in an interview. “We expect Huawei to bring more users and use cases to the community and drive further adoption of Spark in vertical markets”.
Spark, the most active community among Apache projects, has gained strong momentum in the big data market. Huawei’s stance on Spark and in general open source resonated well with the community and the industry. Tushar Shanbhag, VP of Products at Adatao, a fast growing start-up in the Spark ecosystem, expressed the interest and the confidence to collaborate with Huawei. eBay big data team are looking into the possibility to work together with the community and Huawei to bring Kylin OLAP Hadoop project into Spark and boost up Spark OLAP capability.
"The pace of Spark adoption is certainly accelerating," Ben Lorica, Chief Data Scientist at O’Reilly and co-chairman of Strata, well-known in the industry talked at the event. "The strategic positioning of Huawei and its active involvement on Spark community can help expand Spark footprint, while get Huawei exposed to cutting edge technologies. It’s in the good benefit of Spark, Huawei, and the industry as Huawei can help ensure the customer success in Spark implementations”.
Huawei has seen the opportunity to bring Spark into vertical industries, in particular the telecom. “To unlock the full potential of telco data assets, the big data platform needs to be easily plugged into whole IT and network systems, have pre-built capabilities to process and analyze unique telco data like signaling and streaming , while allowing IT, business, data science and network users to continue using their existing skills. Spark provides a solid foundation and framework to pull those pillars off. ” Bing Xiao, the head of big data at Huawei USA, shared Huawei’s vision. “To achieve this, domain specific expertise accumulated in the industry is equally important to glue all moving parts together. The collaboration across the ecosystem among industry and technology vendors is the key to build up this kind of silver bullet solution and make it the reality”.