[
https://issues.apache.org/jira/browse/PHOENIX-2724?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mujtaba Chohan reopened PHOENIX-2724:
-------------------------------------
This is also a performance regression from 4.7. Comparing with current [head of
4.x-HBase-0.98|https://git-wip-us.apache.org/repos/asf?p=phoenix.git;a=commit;h=31534bc169ecfcd03395768aabcb2f3c38fbb8fb]
with
[v4.7.0|https://git-wip-us.apache.org/repos/asf?p=phoenix.git;a=tag;h=abb1410aba958a489e73af2acaa2ac57f3431f6a]
performance with large number of guideposts is significantly slower for point
queries (>15X slower with 80K guideposts in one of the test)
{code}
explain select count(*) from lineitem;
+------------------------------------------------------------------------------------------+
| PLAN
|
+------------------------------------------------------------------------------------------+
| CLIENT 93419-CHUNK 6001026 ROWS 4165407990 BYTES PARALLEL 1-WAY FULL SCAN
OVER LINEITEM |
| SERVER FILTER BY FIRST KEY ONLY
|
| SERVER AGGREGATE INTO SINGLE ROW
|
+------------------------------------------------------------------------------------------+
3 rows selected (8.512 seconds)
select * from lineitem limit 1;
+-------------+------------+------------+---------------+-------------+------------------+-------------+--------+---------------+
| L_ORDERKEY | L_PARTKEY | L_SUPPKEY | L_LINENUMBER | L_QUANTITY |
L_EXTENDEDPRICE | L_DISCOUNT | L_TAX | L_RETURNFLAG |
+-------------+------------+------------+---------------+-------------+------------------+-------------+--------+---------------+
| 1 | 155190 | 7706 | 1 | 17 |
21168.23 | 0.04 | 0.02 | N |
+-------------+------------+------------+---------------+-------------+------------------+-------------+--------+---------------+
1 row selected (8.266 seconds)
Above explain plan and query takes less than 500ms in v4.7.0 running on single
node HBase-0.98.17
{code}
> Query with large number of guideposts is slower compared to no stats
> --------------------------------------------------------------------
>
> Key: PHOENIX-2724
> URL: https://issues.apache.org/jira/browse/PHOENIX-2724
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.7.0
> Environment: Phoenix 4.7.0-RC4, HBase-0.98.17 on a 8 node cluster
> Reporter: Mujtaba Chohan
> Assignee: Samarth Jain
> Fix For: 4.8.0
>
> Attachments: PHOENIX-2724.patch, PHOENIX-2724_addendum.patch,
> PHOENIX-2724_v2.patch
>
>
> With 1MB guidepost width for ~900GB/500M rows table. Queries with short scan
> range gets significantly slower.
> Without stats:
> {code}
> select * from T limit 10; // query execution time <100 msec
> {code}
> With stats:
> {code}
> select * from T limit 10; // query execution time >20 seconds
> Explain plan: CLIENT 876085-CHUNK 476569382 ROWS 876060986727 BYTES SERIAL
> 1-WAY FULL SCAN OVER T SERVER 10 ROW LIMIT CLIENT 10 ROW LIMIT
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
