My personal experience with Tez for last few months was very good. Although I am not from the same company from where the Tez project originally started, the Tez community was very helpful to quickly ramp me up. The community has patiently and promptly gave the constructive feedbacks/comments about my patches. In addition, I attended a few tech-talks and presentations that originally helped me to understand its long term vision.
At the same time, I really appreciate the incubator community to ask these relevant and important questions. Regards, Mohammad Islam On Monday, June 23, 2014 10:52 PM, Rohini Palaniswamy <rohini.adi...@gmail.com> wrote: I will add my two cents in here on the diversity front as a member of the Pig team which has been working closely with Tez team for the past 3 quarters. There has been a whole lot of features and design changes done in Tez driven by requirements from Pig and it has been a pleasure working with the Tez community. We have had a really good collaboration and development has happened at a really really fast pace both in Pig and Tez, with Pig team picking features from Tez snapshot builds as and when they are fixed. We just merged Pig on Tez branch into Pig trunk last month. Most of the collaboration has been through jiras and you can see a lot of linked Pig jiras. Pig on Tez is especially a cross-company team effort in Apache - Yahoo!, Hortonworks, Netflix and LinkedIn and many of the requirements and fixes on Tez has been driven by the security, performance and large scale needs of these companies. http://tinyurl.com/m5tbzxt is the list of Tez jiras created just by the Pig team (48 jiras). There were at least quite a few more created by others when we reported issues in mails. Hive of course is the other big project and Cascading is also into the mix now. With all the widely adopted analytics tools on top of Hadoop adding support for it and many companies adopting it there is very good future for Tez and I think it is just prejudice to call it as a commercial development masquerading as an Apache community. We really see it as a replacement for mapreduce in the near term and pleased with the current results and the future potential as a user and have been investing in the development of Tez as well with dedicated resources. We opened Pig and Hive on Tez for testing for users in Yahoo! in Apr and Cheolsoo opened up Pig on Tez for users in Netflix this month and both Netflix and Y! have plans to get it into production by end of the year and Linkedin also hopes to do the same. With this early adoption from some of the biggest users of hadoop, I am sure Tez will only become more stable and mature in the coming year and gain more traction and wider adoption with other users as well as more companies migrate to Hadoop 2. Regards, Rohini On Mon, Jun 23, 2014 at 9:48 PM, Roman Shaposhnik <r...@apache.org> wrote: > On Mon, Jun 23, 2014 at 9:44 PM, Vinod Kumar Vavilapalli > <vino...@apache.org> wrote: > > I see two threads here > > - Concerns about graduation of Apache Tez project w.r.t diversity based > on the PPMC html page > > - Community related threads about things like whether diversity is a > hard requirement or how > > to grow communities. > > > > AFAICT, Hitesh's latest information on the state of the members of the > team > > (and the updated page) should allay any concerns on diversity. > > > > If people disagree, we can continue this thread along that discussion. > If not, we can close this topic as the graduation discussion/vote proceeds > here and elsewhere. > > > > The other threads may be valuable but OT IMHO and so are better forked > into their own threads. > > I personally think both are very closely related. Like I said, to > me the diversity thread is really in support of the thesis that > the project is capable of growing a community outside of > a forcing function of a single employer. > > That said, I haven't done the due diligence to form my personal > opinion one way or another. I am going to do it this week. > > Thanks, > Roman. >
