HIVE-4095, originally intended to implement,=C2=A0

=
alter table &= lt;target_table_name> exchange partition (<partition spec>) with t= able <source_table_name>;

But= in implementation, =C2=A0Dheeraj Kumar Singh,=C2=A0the original implementor,= =C2=A0se= emed confused and implemented this in inverted manner (target to source).

HIVE-6129 fixed this(in 0.13.0)=C2=A0and now it's consistent to document. B= ut the example has been not working in any version, because HIVE-4095 coerc= ed=C2=A0target_table and=C2=A0source_table should have same partition columns.

https://issues= .apache.org/jira/browse/HIVE-6133 is the needed one for the example to = work. But seemed to be failed to acquire any interest.

Thanks,

Navis

2014-07= -21 11:40 GMT+09:00 Lefty Leverenz <leftyleverenz@gmail.com><= /span>:

I'd be happy to update = the docs, but need some guidance. =C2=A0The sytax confused me originally --= see comments on HIVE-4095= . =C2=A0I'll add this discussion to those comments.

-- Lefty

On Sun, Jul 20, 2014 at 10:22 PM, Andre = Araujo <araujo@pythian.com> wrote:

Indeed! The documentation is a fair bit off.

I've tested the below on Hive 0.12 on CDH and it works fine.
Lefty, would you please update the documentation on the two pages bel= ow?

-----------------------
Source: https://cwiki.apache.or= g/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-ExchangePart= ition

"Exchange Partition" section

{code}=
ALTER TABLE source_table_name EXCHANGE PARTITION (partition_spec= ) WITH TABLE target_table_name;
{code}

This statement lets you move the data in a partition from a table to anothe= r table that has the same schema and partition keys, but does not already h= ave that partition.
For details, see Exchange Partition and HIVE-= 4095.

-----------------------
Source: https://cwiki.apache.org/confluence/display/Hive/Exchange+Parti= tion

The EXCHANGE PARTITION DDL command has been proposed as part of https://= issues.apache.org/jira/browse/HIVE-4095.

The s= yntax is:

{code}
alter table <source_table_name> excha= nge partition (<partition spec>) with table <target_table_name>= ;
{code}

The partition spec can be fully= or partially specified.

The semantics of the above statement is that the data i= s moved from the source table to the target table. Both the tables must hav= e the same schema and the same partition keys. The operation fails in the p= resence of an index.

The partition(s) must exist in the source table and mus NOT exists in = the target one. Consider the following examples:

#= # Full partition spec

{code}

create table T1(a string, b string) partitioned by (ds string);

create table T2(a string, b string) partitioned by (ds string);<= /div>
alter table T1 add partition (ds =3D '1');
{cod= e}

The operation:

{code}<= /div>

alter table T1 exchange partition (ds=3D'1') with table T2;
{code}

moves the data from T1 to T2@= ds=3D1. The operation fails if T2@ds=3D1 already exists or T1 and T2 have d= ifferent schemas and/or partition keys.

## Partial partition spec

{cod= e}
create table T1(a string, b string) partitioned by (ds string,= hr string);
create table T2(a string, b string) partitioned by (= ds string, hr string);

alter table T1 add partition (ds =3D '1', hr =3D '00')= ;
alter table T1 add partition (ds =3D '1', hr =3D '0= 1');
alter table T1 add partition (ds =3D '1', hr =3D= '03');

{code}

The operation:

{code}
alter table T1 exchange partition (ds=3D'1'= ) with table T2;
{code}

moves the 3 partitions from T1 to T2. The operation fails if any of the par= titions already exist on T2 or if T1 and T2 have different schemas and/or p= artition keys.

Either all the partitions of T1 will get created or the whole operatio= n will fail. All partitions of T1 are dropped.

On 21 July 2014 05:52, Kristof Vanbecelaere <kristof.vanbec= elaere@gmail.com> wrote:

I think the documentation r= elated to exchanging partitions is not accurate

https://cwiki.apache.org/confluence/display/Hive/Exchange+P= artition

when I try it out on hortonworks sandbox 2.1 which runs= Hive 0.13 I get this:

hive> create table T1(a string, b string) = partitioned by (ds string);

OK

Time taken: 0.72 s= econds

hive> create ta= ble T2(a string, b string);

OK

Time taken: 0.357 = seconds

hive> alter tab= le T1 exchange partition (ds=3D'1') with table T2;

FAILED: SemanticEx= ception [Error 10235]: Tables have incompatible schemas and their partition= s=C2=A0 cannot be exchanged.

--
=
Andr=C3=A9 Ara=C3=BAjo
Big Data Consultant/Solutions Ar= chitect
The Pythian Group - Australia - www.pythian.com

Office (calls from within Australia): 1300 366 021 x1270
Office= (international): +61 2 8016 7000 =C2=A0x270 OR +1 613 565 8696=C2=A0=C2=A0 x1270
Mobile: +61 410 323 559
Fax: +61 2 9805 0544
IM: = pythianaraujo @ AIM/MSN/Y! or araujo@pythian.com @ GTalk

=E2=80=9CSuccess is not about standing at the=C2=A0top, it's the steps you=C2=A0leave behind.= =E2=80=9D =E2=80=94=C2=A0Iker Pou (rock climber)

--

--089e015373cc0324a204feac2a5f--