Hi Stephen, Yes --- this week Daniel and I did benchmark testing of SparkGraphComputer on a cluster over Friendster (2.5 billion edges). There were so many little "knick nack" things we discovered. Not so much bugs, but optimizations that are crucial at large scale. With that said, I think another week of a "code freeze" would be good. I'd like to wrap up our benchmark tomorrow and present to everyone our findings. Some of the things were learned were huge and so beneficial.
If everyone else is cool with another week of code freeze that is great. If not, I will be done with our benchmark work tomorrow and all closed up on the code by COB tomorrow. Thus, all good for a release Monday. Up to everyone else. Thanks, Marko. http://markorodriguez.com On Jan 28, 2016, at 6:10 PM, Stephen Mallette <spmalle...@gmail.com> wrote: > Code freeze kinda didn't work so well this week, in the sense that we found > some bugs and other odds and ends during testing and we had to push some > commits through. Of course, that's part of what this code freeze is for - > we take a moment to do some more detailed testing on a release before we > pull the trigger. > > So that said, I'm wondering if we shouldn't take another "code freeze" for > another week just to be sure everything is stable and good to go. > Thoughts? > > On Mon, Jan 25, 2016 at 9:09 AM, Stephen Mallette <spmalle...@gmail.com> > wrote: > >> Just a reminder that we're officially freezing the repo starting today. >> Just documentation changes are allowed at this point. Please review upgrade >> documentation/changelog and update as needed. Also, please find some time >> to test and yell if you run into problems. We'll prepare for release VOTE >> next Monday. There are still some open issues hanging out there in JIRA. >> If your name is attached to those can you please update those so that we >> can either close them or move them off to another version. Thanks! >> >> On Fri, Jan 22, 2016 at 11:50 AM, Stephen Mallette <spmalle...@gmail.com> >> wrote: >> >>> Can committers please do some reviews and throw in some votes so that we >>> can get these final PRs merged in? >>> >>> https://github.com/apache/incubator-tinkerpop/pull/205 >>> https://github.com/apache/incubator-tinkerpop/pull/201 >>> https://github.com/apache/incubator-tinkerpop/pull/170 (marko - you said >>> you were working with kuppitz on this one some time back) >>> >>> I guess we don't need to worry about these two for right now: >>> >>> https://github.com/apache/incubator-tinkerpop/pull/186 >>> https://github.com/apache/incubator-tinkerpop/pull/195 >>> >>> thanks, >>> >>> Stephen >>> >>> >>> On Thu, Jan 21, 2016 at 6:22 AM, Stephen Mallette <spmalle...@gmail.com> >>> wrote: >>> >>>> I cleaned up the issue list based on feedback. >>>> >>>> Jason, any update on this one: >>>> >>>> https://issues.apache.org/jira/browse/TINKERPOP-964 >>>> >>>> can we close? or do we need to move forward to next version? >>>> >>>> >>>> >>>> On Tue, Jan 19, 2016 at 7:23 PM, Marko Rodriguez <okramma...@gmail.com> >>>> wrote: >>>> >>>>> Hi, >>>>> >>>>> Yea. Kuppitz and I are testing SparkGraphComputer on a cluster (both >>>>> SparkServer and Hadoop2). We are having problems with >>>>> ClassNotFound/jar-style exceptions. Not really ticket worthy as its not >>>>> "code" as much as us just learning about the patterns people should use >>>>> for >>>>> deploying jars. We will have this all settled and tested at scale (Enron >>>>> dataset) by the end of the week. >>>>> >>>>> Thanks, >>>>> Marko. >>>>> >>>>> http://markorodriguez.com >>>>> >>>>> On Jan 19, 2016, at 5:04 PM, Daniel Kuppitz <m...@gremlin.guru> wrote: >>>>> >>>>>> Forgot to mention: Marko and I are still trying to fix some >>>>>> SparkGraphComputer issues (I don't think there're open tickets for the >>>>>> stuff we're doing). >>>>>> >>>>>> Cheers, >>>>>> Daniel >>>>>> >>>>>> >>>>>> On Wed, Jan 20, 2016 at 1:02 AM, Daniel Kuppitz <m...@gremlin.guru> >>>>> wrote: >>>>>> >>>>>>> I don't think I get the script things done (TINKERPOP-927 >>>>>>> <https://issues.apache.org/jira/browse/TINKERPOP-927> and >>>>> TINKERPOP-986 >>>>>>> <https://issues.apache.org/jira/browse/TINKERPOP-986>). I'm waiting >>>>> for >>>>>>> feedback from Michael for TINKERPOP-939 >>>>>>> <https://issues.apache.org/jira/browse/TINKERPOP-939> (not much >>>>>>> confidence here either). >>>>>>> TINKERPOP-943 <https://issues.apache.org/jira/browse/TINKERPOP-943> >>>>> has >>>>>>> an open PR (simple stuff, will def. make it into 3.1.1). >>>>>>> TINKERPOP-818 <https://issues.apache.org/jira/browse/TINKERPOP-818> >>>>>>> should be easy, chances are high that you'll see a PR in the coming >>>>> days. >>>>>>> >>>>>>> Cheers, >>>>>>> Daniel >>>>>>> >>>>>>> >>>>>>> On Tue, Jan 19, 2016 at 9:31 PM, Stephen Mallette < >>>>> spmalle...@gmail.com> >>>>>>> wrote: >>>>>>> >>>>>>>> We have the rest of this week to hack away until code freeze >>>>> arrives next >>>>>>>> Monday, January 25th 2016. As it stands we have a handful of items >>>>>>>> remaining - some are already in pull requests awaiting review. >>>>>>>> >>>>>>>> Anyone foresee any troubles getting their bits done in time for code >>>>>>>> freeze? >>>>>>>> >>>>>>>> We also have these open issues which are unassigned to anyone: >>>>>>>> >>>>>>>> + Develop a less error prone way for rewriting strategies - >>>>>>>> https://issues.apache.org/jira/browse/TINKERPOP-882 - don't >>>>> imagine we >>>>>>>> will >>>>>>>> make this happen. >>>>>>>> + Graph Configuration Class - >>>>>>>> https://issues.apache.org/jira/browse/TINKERPOP-659 >>>>>>>> + StructureStandardTestSuite has file I/O issues on Windows >>>>>>>> https://issues.apache.org/jira/browse/TINKERPOP-1041 - Jason, >>>>> sorry to >>>>>>>> make >>>>>>>> you "Windows guy" on this one, but is this one you can easily solve? >>>>>>>> + Validate dependency grabs that have TinkerPop dependencies - >>>>>>>> https://issues.apache.org/jira/browse/TINKERPOP-893 - this was >>>>>>>> low-hanging >>>>>>>> fruit that we thought someone in the community might pick up - we >>>>> can >>>>>>>> probably push that off to another version. >>>>>>>> >>>>>>>> Comments? >>>>>>>> >>>>>>> >>>>>>> >>>>> >>>>> >>>> >>> >>
