Cool. Got it. On Nov 17, 2012 1:40 PM, "Miki Tebeka" <[email protected]> wrote:
> I mean the Python tools (easy_install avro). > > > > On Sat, Nov 17, 2012 at 7:46 AM, ranjith raghunath < > [email protected]> wrote: > >> Thanks for response. When you say avro tools you mean avro-tools-.....jar >> right? >> >> Let me also run the flow by all of you. Use sqoop to download data from >> an rdbms to avro format. Use avro tools to extract schema file. Use avro >> serde to generate/update hive table. So this would eliminate the need for >> statically mapping the fields in hive. >> >> Is this flow one that makes sense? >> On Nov 17, 2012 9:36 AM, "Miki Tebeka" <[email protected]> wrote: >> >>> You can use the "avro" utility that comes when you install the Python >>> package (or fastavro if you need 3.X support). Then run "avro cat >>> --print-schema /path/to/avro/file". >>> >>> >>> On Sat, Nov 17, 2012 at 5:41 AM, ranjith raghunath < >>> [email protected]> wrote: >>> >>>> I could really use some advice on this topic. >>>> >>>> I am pulling files in avro format from an external source (outside of >>>> the cluster). How can I generate the avro schema file? The end goal is to >>>> have it exposed in Hive. >>>> >>>> Thanks, >>>> Ranjith >>>> >>> >>> >
