Hi Luke and Alexander. Thanks for asking question and here is reason I did list storage engine.
S2Graph has been used HBase as primary storage engine. I think there is no reason we need to change this. However, I also think there is no reason we should only support HBase. We realized that lots of codes can be independent to storage backend, so we abstract away storage dependent codes at S2GRAPH-17. after this refactoring, it becomes easy for others who want to use different storage other than HBase to connect to their choice for storage. Personally I think it would be better to give user more options. For example, http://thinkaurelius.github.io/titan/ support various storages(Cassandra, HBase, BerkeleyDB, but seems primary is Casssandra) and I think S2Graph can also support these options, but primary is HBase. What others think about this? Also regarding Query Graphical User Interface, I have no idea what other existing project can be used. If it is possible to re-use existing projects, then I prefer to use them. Please guide me what are these existing projects(I would love to try Zeppelin though). Thanks. Doyung Yoon On Thu, Mar 24, 2016 at 8:54 AM Luke Han <[email protected]> wrote: > Hi > For Storage Engines, are we trying to extend to others rather than > HBase? > > Thanks. > Luke > > > Best Regards! > --------------------- > > Luke Han > > On Wed, Mar 23, 2016 at 10:31 PM, Kim, Min-Seok <[email protected]> > wrote: > > > updated and added > > > > > > - > > > > Batch Jobs(S2Lambda) > > - > > > > Kafka to HDFS (WAL) > > - > > > > Streaming Cooccurrence across labels(user-user/item-item > similarity) > > - > > > > OLAP operations on (WAL or KAFKA) > > - > > > > A/B Testing capabilities > > - Multi-armed Bandit to select the best query > > > > > > I think A/B and MAB can be component themselves, they could be merged > into > > other components. > > > > Thanks > > > > > > > > > > > > 2016년 3월 22일 (화) 오전 7:57, DO YUNG YOON <[email protected]>님이 작성: > > > > > Hi folks. > > > > > > I just want to open up discussion on our project roadmap. > > > > > > > > > > > > https://docs.google.com/document/d/1QSEf628QHrLmky16cJN_wIv0H_cfi1E9NGLqKMpdsNE/edit?usp=sharing > > > > > > > > > Things I wrote on link is completely draft(I just list up all I can > think > > > of now), please feel free to change as it is necessary. > > > Some of them might easy and some of them would take time, so I also > want > > to > > > ask what others think about first release. > > > Personally, I think It would be great if we can list up our load map, > > then > > > decide priority, and talks about our first release. > > > Also, I think once we decide road map, then this road map can be used > as > > > component at JIRA. currently there is no component so it is hard to > guess > > > what each issue is about. > > > > > > Looking forward to hear what others think. > > > > > >
