Hi Denis, Interesting results, thank you.
Unfortunately Ignite Persistence came twice in used features. And it becomes a a little bit unclear which measurement should be considered correct: max, min, or average. Sincerely, Dmitriy Pavlov чт, 22 февр. 2018 г. в 4:17, Dmitriy Setrakyan <dsetrak...@apache.org>: > Very useful. > > BTW, here is the ticket for MultiMap implementation. It has a very nice > design proposal and looks fairly simple to implement. > https://issues.apache.org/jira/browse/IGNITE-640 > > Hope someone in the community can pick it up. > > D. > > On Wed, Feb 21, 2018 at 3:56 PM, Denis Magda <dma...@apache.org> wrote: > > > Igniters, > > > > The survey is off and here is a snapshot of the results: > > https://www.surveymonkey.com/results/SM-2T8S87DH8/ > > > > Noteworthy observations: > > > > - 87% of respondents are software architects and engineers. These are > > the folks who help us to drive the project! > > > > > > - Most of Ignite use cases are in Banks and HighTech - this is what > the > > consistency and in-memory speed are valued for! > > > > > > - Majority of the deployments store up to 100 GB of data in up to 20 > > nodes clusters while 10% of deployments maintain hundreds and > thousands > > nodes clusters - getting ready, community! > > > > > > - Ignite persistence is getting as valuable as the compute grid. > > > > Pay attention to the last question that contains a written individual > > feedback. It should guide our plans. For instance, I like this proposal: > > > > MultiMap support like in Guava is badly needed. Hazelcast has this and is > > highly used collection type, so much so this forces us to have to use > > Hazelcast over Ignite sadly in many apps > > > > -- > > Denis > > >
