Hello Joo Yeon. Thanks for being our first release manager.
I started the implementation of (2) RDD.cache() today after the PR #94 ( https://github.com/apache/incubator-nemo/pull/94), which was a blocking issue for caching, is merged. In my expectation, a PR support basic persist() and cache() features (specifically, excluding persisting on both disk & memory, off-heap memory, and replication) before Aug. 16st. Best, Sanha 2018-08-09 15:39 GMT+09:00 John Yang <[email protected]>: > Hi Joo Yeon, > > I'll be working on (1) and (3). It turns out (1) depends on (4) for > IR-based dynamic optimizations, and (3) depends on (2) for caching ALS > models. So I'll probably need maximum time until Aug. 21st. > > While waiting for (4) and (2) to be resolved, I'm hoping to file a PR in > 1~2 days that partially addresses (1) by resolving NEMO-178( > https://issues.apache.org/jira/browse/NEMO-178). > > Thanks, > John > > > On Thu, Aug 9, 2018 at 3:24 PM Joo Yeon Kim <[email protected]> wrote: > > > Hi all. > > > > I would like to start a list of features that we are currently working > on, > > and wish to be included in the first release. > > > > The below 4 items provide a starting point: > > > > 1. Speculative execution using task cloning > > 2. Support for RDD.cache() > > 3. Support for Spark ALS > > 4. Generalized dynamic optimization > > > > @To all the committers, please reply to this mail thread > > a) if you are in charge of one of the 4 features above, with which > feature > > you will work on and the expected timeline. > > b) if there is a feature you would like to include in this release. > > > > We expect to freeze our code around Aug. 21st, so that the release can be > > made by the end of August. > > > > Once I receive all the emails, I'll send out another email with the JIRA > > issue links and the schedule. > > > > Happy coding! > > > > Best, > > Joo Yeon > > >
