Last week Stumbleupon hosted the eighth HBase User Group (HUG8) meetup at our office in downtown San Francisco. Around 90 HBase enthusiasts gathered to attend the presentations and share their stories. We’ve hosted a previous meetup in the past—and this one was by far the largest.
Jean-Daniel Cryans, our awesome StumbleUpon Database Engineer was in charge of organizing the event and has kindly put together an overview (below) of the presentations, along with links to the pdfs. Thank you, JD – you rock!
Thanks to everyone who came! Please stay tuned, we will be posting more footage from the event in the following days on our developers blog.
The StumbleUpon Team
Jean-Daniel’s presentation was a small introduction to HBase followed by a summary of 2009, during which two major versions were released and three new committers were added to the team. He also talked about the latest release, 0.20.3, which is available for download here, and the upcoming 0.21. (Download presentation)
Ryan Rawson, System Architect at StumbleUpon and also HBase committer, presented how we were able to achieve 1.1M inserts per second by using multi-threading inside the HBase’s client library. (Download presentation)
Adam Silberstein, Research Scientist at Yahoo!, talked about his team’s research in the domain of large-scale distributed databases benchmarking and showed how HBase compared to other systems according to their methodology. (Download presentation)
Bradford Stephens, co-founder at Drawn to Scale, gave the last presentation on how HBase can be integrated in the search ecosystem; followed by an example on how they use it at his company. (Download presentation)
Adam Silberstein from Yahoo! talks about setting up HBase.