phoenix-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chunhui Liu <leeyc...@gmail.com>
Subject Re: Row value constructors failed on the index, when len(table's pks) > 2 and table's 1st pk is index's last pk
Date Thu, 15 Oct 2015 07:36:31 GMT
Test with patch for PHOENIX-2319, issue still happened.

When PHOENIX-2319 was triggered, no index upsert into hbase.

For this issue, UPSERT seems ok, "SELECT * FROM IDX_T" works fine.

Thanks,
Chunhui

2015-10-15 14:26 GMT+08:00 James Taylor <jamestaylor@apache.org>:

> Any difference if you apply the patch for PHOENIX-2319?
> Thanks,
> James
>
> On Wed, Oct 14, 2015 at 10:15 PM, Chunhui Liu <leeychee@gmail.com> wrote:
>
>> Hi team,
>>
>> When I try to use paged query on secondary index, I found a issue.
>>
>> 1. Table has more than 2 primary keys;
>> 2. Table's 1st pk as index's last pk; eg. table's pks are (pk1, pk2,
>> pk3), the failed index's pks are (pk2, pk3, pk1); table's pks are (1, 2, 3,
>> 4), failed index's pks are (2, 3, 4, 5, 1);
>> 3. Use row value constructors on index with another condition that use
>> one pks(not the table's 1st pk);
>> 4. You will get "DEGENERATE SCAN OVER TABLE_NAME"
>>
>> Here is the Test SQL
>> ---------------------------------------------------------------------
>> DROP TABLE IF EXISTS T;
>>
>> CREATE TABLE IF NOT EXISTS T (
>>    PK1 VARCHAR not null,
>>    PK2 VARCHAR not null,
>>    PK3 VARCHAR not null,
>>    V1  VARCHAR,
>>    CONSTRAINT PK PRIMARY KEY (PK1, PK2, PK3)
>> );
>>
>> CREATE INDEX IDX_T ON T
>> (
>>    PK2, PK3, PK1
>> );
>>
>> UPSERT INTO T VALUES('100', '200', '300', 'V');
>> UPSERT INTO T VALUES('101', '201', '301', 'V');
>> UPSERT INTO T VALUES('102', '202', '302', 'V');
>> UPSERT INTO T VALUES('103', '203', '303', 'V');
>> UPSERT INTO T VALUES('104', '204', '304', 'V');
>>
>> SELECT * FROM T;
>>
>> EXPLAIN
>> SELECT PK1, PK2, PK3 FROM T WHERE
>> (PK2, PK3, PK1) >= ('202', '302', '102')
>> AND PK2 < '204'
>> LIMIT 10;
>> ---------------------------------------------------------------------
>>
>> I've tried 3 primary key, here is the results.
>> 1. table's pks are (pk1, pk2, pk3);
>> 2. 132 means (pk1, pk3, pk2);
>>
>> | index's pks order     | result                               |
>> | --------------------- | ------------------------------------ |
>> | 132                   | correct                              |
>> | 213                   | correct                              |
>> | 231                   | fail                                 |
>> | 312                   | correct                              |
>> | 321                   | correct                              |
>>
>> I've also test this on table with 4, 5 pks
>> | len(pks)              | failed order                         |
>> | --------------------- | ------------------------------------ |
>> | 3                     | 231                                  |
>> | 4                     | 2341                                 |
>> | 5                     | 23451                                |
>>
>>
>> Thanks,
>> Chunhui
>>
>>
>

Mime
View raw message