I was able to find the property with some digging around and
experimentation. Never knew that ppd had something to do with this
property.
On Thu, Apr 3, 2014 at 7:23 PM, Stephen Sprague wrote:
> wow. good find. i hope these config settings are well documented and that
> you didn't have to spend
It is not essentially the client side. The file can be generated by a
running MR job or any process for that matter which can then feed data to
the Hadoop cluster to run hive queries.
-Abhay
On Thu, Apr 3, 2014 at 7:40 PM, Bogala, Chandra Reddy wrote:
> I thought ORC file can be generated only
Thanks again Petter, the custom input format was exactly what I needed.
Here is example of my code in case anyone is interested
https://github.com/quicklyNotQuigley/nest
Basically gives you SQL access to arbitrary json data. I know there are
solutions for dealing with JSON data in hive fields but
Just to add to my previous question, I see this example in Hive
documentation for the "reflect" UDF. What object or string is the "isEmpty"
method working on in the example below? What I am trying to do with
hashCode() is something similar.
SELECT reflect("java.lang.String", "valueOf", 1),
Hi,
I'm new to using Parquet and ORC files and I'm hitting a problem with
querying nested data. Can those files formats be used to query deeply
nested data?
If yes, why I am getting an error with the SerDes for both of them?
Here's the background:
I'm starting from a JSON data file like this:
It's probably not as pretty as the new built-in version, but this
allows scripted UDFs in any javax.script language:
https://github.com/livingsocial/HiveSwarm/blob/master/src/main/java/com/livingsocial/hive/udf/ScriptedUDF.java
On Thu, Apr 3, 2014 at 4:35 PM, Andy Srine wrote:
> Thanks Edward. Bu
Thanks Edward. But "inline groovy" is available on Hive 13 right? I am
using an older version.
Best,
Andy
On Thu, Apr 3, 2014 at 11:37 AM, Edward Capriolo wrote:
> You can write UDF's in groovy now. That pretty much means. You can just
> write a quick method inline now. Makese udf reflect much
You can write UDF's in groovy now. That pretty much means. You can just
write a quick method inline now. Makese udf reflect much less useful.
On Thu, Apr 3, 2014 at 2:22 PM, Andy Srine wrote:
> Thanks Szehon and Peyman, I want to call hashCode() on the UUID object.
> This object is stored in th
Thanks Szehon and Peyman, I want to call hashCode() on the UUID object.
This object is stored in the table as a string, but I can convert it to
UUID. Thats not the problem.
Basically the question is, how do we call this reflect UDF on methods that
takes no arguments? How do I do the following usin
I thought ORC file can be generated only by running hive query on staging table
and inserting into ORC table. If there is option to generate ORC file at client
side by using java code then can you share that code or links related to that?
Thanks,
Chandra
From: Abhay Bansal [mailto:abhaybansal.1.
wow. good find. i hope these config settings are well documented and that
you didn't have to spend alot time searching for that. Interesting that
the default isn't true for this one.
On Wed, Apr 2, 2014 at 11:00 PM, Abhay Bansal wrote:
> I was able to resolve the issue by setting "hive.optimize
Maybe your intention is the following:
reflect("java.util.UUID", "randomUUID")
On Thu, Apr 3, 2014 at 2:33 AM, Szehon Ho wrote:
> Hi, according to the description of the reflect UDF, you are trying to
> call java.util.UUID.hashcode(uidString), which doesnt seem to be an
> existing method on eit
12 matches
Mail list logo
