Hi Talat, In all honesty I don't have the same time I used to, to look into this. I have been experimenting using Arrow with multi-dimensional array-based data but nothing else. I would therefore be learning probably as much as you if this project was to go ahead. Lewis
On Thu, Mar 15, 2018 at 3:46 PM, Talat Uyarer <ta...@uyarer.com> wrote: > @Lewis I found a PR[0] on Arrow Git repo. I guess they stuck with avro-c > library. Do you know do they need implement same thing for all languages > which are supported by them or they just need to implement a wrapper ? > > If we can use Arrow for our internal serialization, Gora will be super > fast with zero copy support. :) > > [0] https://github.com/apache/arrow/pull/1026 > > My 2 cent > > On Thu, Mar 15, 2018 at 12:24 AM, lewis john mcgibbney <lewi...@apache.org > > wrote: > >> Hi Renato, >> >> On Wed, Mar 14, 2018 at 3:22 PM, Renato Marroquín Mogrovejo < >> renatoj.marroq...@gmail.com> wrote: >> >>> Hey guys, >>> >>> There might not be an integration/convertors of Arrow to Avro (and/or >>> viceversa) because there are parquet readers that can take avro and once >>> stuff is in parquet, then arrow can be used directly. >>> >> >> Yes there might not be. I actually raised this issue [0] a wee while ago >> on the Arrow list. At that time I was told, "...The use case you outline >> makes a lot of sense for Arrow to help out with. We don't yet have an AVRO >> <> Arrow converter written but it is something that would be great to >> have." So maybe that would be something to keep in mind. >> >> [0] https://s.apache.org/2GwS >> >> >>> Regarding if an integration of Parquet with Gora, I think it would be >>> interesting to make it easier for people to read and write parquet files by >>> providing a higher level api as Gora provides. However, for you @Talat, >>> that knows Gora pretty well, maybe you could take another project that >>> helps Gora more. For example, fixing the integration with Nutch. There are >>> multiple loose ends in Nutch 2.x and Gora that we have neglected as a >>> community. >>> IMHO that should be GSOC project. >>> >> >> ACK, other existing projects which consume Gora are (off the top of my >> head), >> >> - Chukwa - https://s.apache.org/cW6a >> - Giraph - https://github.com/apache/giraph/tree/trunk/giraph-gora >> - Camel - https://camel.apache.org/gora.html >> - Nutch 2.X - https://github.com/apache/nutch/tree/2.x >> >> An interesting idea I had where Gora could be implemented would be in >> Hadoop metrics >> >> https://hadoop.apache.org/docs/current/hadoop-project-dist/ >> hadoop-common/Metrics.html >> >> This would provide provide a text book usage for Gora to store Hadoop >> metrics in some datastore which would then be exposed for query and >> analysis. >> >>> I can't mentored it because I do not have enough insights on this, but >>> @Lewis and @Talat you can probably tackle this as mentor and student. This >>> would be an awesome contribution to the project as there are quite a lot of >>> people going over Nutch and trying to use it with Gora. >>> Just my 2c >>> >>> >> Understood Renato, no biggie. Thanks for your input. I know you are >> working with Parquet alot these days so your input is appreciated. >> Lewis >> > > > > -- > Talat UYARER > Websitesi: http://talat.uyarer.com > Twitter: http://twitter.com/talatuyarer > Linkedin: http://tr.linkedin.com/pub/talat-uyarer/10/142/304 > -- http://home.apache.org/~lewismc/ http://people.apache.org/keys/committer/lewismc
