Apache Phoenix Performance Result :: Mon Jun 20 19:41:11 PDT 2016

Test Suite: STANDARD

Queries by Category



Aggregation

select count(*) from AGGREGATION_TABLE
1s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.41s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from AGGREGATION_TABLE group by HOST
1.1s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.49s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, count(1) from AGGREGATION_TABLE group by CORE
0.98s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.53s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, host, sum(ACTIVE_VISITOR) from AGGREGATION_TABLE group by core, host
2.36s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.41s 4.X-HBASE-0.98-0.98.17-HADOOP2

select AVG(DB), MIN(CORE), MAX(ACTIVE_VISITOR) from AGGREGATION_TABLE where host='CS' and DOMAIN<>'Google.com'
1.05s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.96s 4.X-HBASE-0.98-0.98.17-HADOOP2

select round(date, 'day',50), sum(core)+2 from AGGREGATION_TABLE where (CORE<10 or CORE>90) group by round(date, 'day',50), host
1.19s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.3s 4.X-HBASE-0.98-0.98.17-HADOOP2

select trunc(date, 'hour', 1000), count(DB), count(ACTIVE_VISITOR) from AGGREGATION_TABLE where feature='Login' and (core<50 or db<20 or active_visitor > 200) group by trunc(date, 'hour', 1000)
2.24s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.34s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(active_visitor) count from AGGREGATION_TABLE
2.08s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.08s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(distinct active_visitor) count from AGGREGATION_TABLE
2.78s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.84s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core), count(db), count(active_visitor) from AGGREGATION_TABLE
2.43s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.5s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, count(db) count from AGGREGATION_TABLE group by CORE
0.97s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.85s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core < 10 and db < 200
1.19s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.36s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core < 10 and active_visitor < 1000
2.03s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.06s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core > 90
1.09s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.12s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from WIDE_PK
3.13s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.44s 4.X-HBASE-0.98-0.98.17-HADOOP2


Aggregation with filter no non-pk

select count(*) from TABLE_6CF where f < 10000
1.76s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.72s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF where f < 10000
1.61s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.4s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from WIDE_PK where column2<10
2.53s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.5s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from WIDE_PK where column2<50 and column3<50 and column4<50 and column5<50
2.83s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.27s 4.X-HBASE-0.98-0.98.17-HADOOP2


ColumnFamilies-Single

select count(1) from TABLE_6CF where K1 in ('A', 'B', 'C', 'D')
0.98s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.04s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from TABLE_6CF where K1 in ('A', 'C')
0.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.62s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_6CF where f < 1000
0.92s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.64s 4.X-HBASE-0.98-0.98.17-HADOOP2

select a,b,c,d,e,f from TABLE_6CF where B>1000 and B<2000 and f>1000 and f<2000
1.82s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.64s 4.X-HBASE-0.98-0.98.17-HADOOP2

select sum(a),sum(b),sum(c),sum(d) from TABLE_6CF group by K1
5.01s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
4.69s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from TABLE_1CF where K1 in ('A', 'B', 'C', 'D')
1.51s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.39s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from TABLE_1CF where K1 in ('A', 'C')
1.1s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.74s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF where f < 1000
1.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.32s 4.X-HBASE-0.98-0.98.17-HADOOP2

select a,b,c,d,e,f from TABLE_1CF where B>1000 and B<2000 and f>1000 and f<2000
2.11s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.51s 4.X-HBASE-0.98-0.98.17-HADOOP2

select sum(a),sum(b),sum(c),sum(d) from TABLE_1CF group by K1
1.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.92s 4.X-HBASE-0.98-0.98.17-HADOOP2


Count Distinct

select count(distinct core), count(distinct db), count(distinct active_visitor) from AGGREGATION_TABLE
3.51s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.68s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, count(distinct db) count from AGGREGATION_TABLE group by CORE
2.59s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.55s 4.X-HBASE-0.98-0.98.17-HADOOP2


IN/LIKE Clause

select count(core) from AGGREGATION_TABLE where core IN (1,25,50,75,100)
1.04s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.13s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from IN_LIKE_TABLE where K1 in ('Z', 'D', 'F', 'X')
0.25s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.37s 4.X-HBASE-0.98-0.98.17-HADOOP2

select * from IN_LIKE_TABLE where K1 in ('A','B') and K2 in ('xxx1000000','xxx2000000','xxx2000001')
0.01s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0s 4.X-HBASE-0.98-0.98.17-HADOOP2

select * from IN_LIKE_TABLE where K1 in ('A', 'C', 'E', 'F', 'H') and K2 like 'xxxxxx1%'
0.02s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.01s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from IN_LIKE_TABLE where K1 in ('X','A','B','C','D','E','F','G','H','I','Z') and K2 like '%1%'
1.75s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.3s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from IN_LIKE_TABLE where (K1 in ('A','E','F','G','H','I','Z') or K1 like 'E%') and (K2 like '%xx1%' or K2 like '%xxxxxx2%')
1.21s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.67s 4.X-HBASE-0.98-0.98.17-HADOOP2


Index

select count(*) from INDEXED_TABLE
0.52s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.52s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(*) from INDEXED_TABLE
0.49s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.5s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(*) from INDEXED_TABLE
0.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.85s 4.X-HBASE-0.98-0.98.17-HADOOP2

select round(date, 'day',50), sum(core)+2 from INDEXED_TABLE where (CORE<10 or CORE>90) group by round(date, 'day',50), host
0.27s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.26s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ round(date, 'day',50), sum(core)+2 from INDEXED_TABLE where (CORE<10 or CORE>90) group by round(date, 'day',50), host
0.95s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.04s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core IN (1,100)
0.04s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.05s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(core) from INDEXED_TABLE where core IN (1,100)
0.03s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.04s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core IN (1,100)
0.7s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.81s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core IN (1,25,50,75,100)
0.07s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.07s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(core) from INDEXED_TABLE where core IN (1,25,50,75,100)
0.06s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.06s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core IN (1,25,50,75,100)
0.68s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.82s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core < 10 and db < 200
0.12s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.13s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(core) from INDEXED_TABLE where core < 10 and db < 200
11.09s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
11.1s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core < 10 and db < 200
0.83s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.97s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core < 100 and db < 2000
1.18s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.54s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core < 100 and db < 2000
1.16s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.39s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core > 90
0.07s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.08s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core > 90
0.73s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.83s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core > 0
0.59s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.62s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core > 0
0.85s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.97s 4.X-HBASE-0.98-0.98.17-HADOOP2

select db from INDEXED_TABLE where db >5 and db <7 order by db
0.04s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.03s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ db from INDEXED_TABLE where db >5 and db <7 order by db
1.32s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.02s 4.X-HBASE-0.98-0.98.17-HADOOP2

select db from INDEXED_TABLE order by db limit 10
0.05s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.02s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ db from INDEXED_TABLE order by db limit 10
1.36s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.35s 4.X-HBASE-0.98-0.98.17-HADOOP2


Index Creation After Data

CREATE INDEX idx5 ON INDEXED_TABLE (CORE) INCLUDE (DB,ACTIVE_VISITOR)
37.51s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
26.24s 4.X-HBASE-0.98-0.98.17-HADOOP2


LIMIT

select * from AGGREGATION_TABLE LIMIT 100
0.02s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.02s 4.X-HBASE-0.98-0.98.17-HADOOP2


Large Group By

select count(*) from TABLE_6CF group by A
0.72s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.93s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_6CF group by B
2.26s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.86s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF group by A
1.57s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.31s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF group by B
2.99s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.06s 4.X-HBASE-0.98-0.98.17-HADOOP2


Load Data

LOAD DATA AGGREGATION_TABLE [2000000 ROWS]
152.17s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
140.49s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA IN_LIKE_TABLE [2000000 ROWS]
106.82s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
104.25s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA INDEXED_TABLE [2000000 ROWS]
383.92s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
367.53s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA TABLE_6CF [2000000 ROWS]
145.43s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
142.44s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA SALTED_TABLE [2000000 ROWS]
141.73s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
143.13s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA TABLE_1CF [2000000 ROWS]
126.44s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
128.48s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA WIDE_PK [2000000 ROWS]
284.28s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
309.02s 4.X-HBASE-0.98-0.98.17-HADOOP2


Percentile

select percentile_cont(0.9) within group (order by active_visitor asc) from AGGREGATION_TABLE
2.75s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.87s 4.X-HBASE-0.98-0.98.17-HADOOP2

select percentile_cont(0.1) within group (order by active_visitor desc) from AGGREGATION_TABLE
2.75s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.85s 4.X-HBASE-0.98-0.98.17-HADOOP2

select percentile_cont(0.5) within group (order by core asc), percentile_cont(0.5) within group (order by db asc), percentile_cont(0.5) within group (order by active_visitor asc) from AGGREGATION_TABLE
3.53s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.74s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, percentile_cont(0.9999) within group (order by db asc) from AGGREGATION_TABLE group by CORE
2.67s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.67s 4.X-HBASE-0.98-0.98.17-HADOOP2


Reverse Scan

select * from WIDE_PK order by mypk DESC limit 10
0.12s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.43s 4.X-HBASE-0.98-0.98.17-HADOOP2


Salting

select count(1) from SALTED_TABLE where K1 in ('A', 'B', 'C', 'D')
1.37s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.89s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from SALTED_TABLE where D < 1000
0.68s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.77s 4.X-HBASE-0.98-0.98.17-HADOOP2

select avg(c),avg(d) from SALTED_TABLE group by K1
2.9s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.93s 4.X-HBASE-0.98-0.98.17-HADOOP2




Queries by Table



AGGREGATION_TABLE


Table used for various aggregation functions

ROWS 2000000 [2M]

DDL CREATE TABLE $TABLE (HOST CHAR(2) NOT NULL,DOMAIN VARCHAR NOT NULL,FEATURE VARCHAR NOT NULL,DATE DATE NOT NULL,USAGE.CORE BIGINT,USAGE.DB BIGINT,STATS.ACTIVE_VISITOR INTEGER CONSTRAINT PK PRIMARY KEY (HOST, DOMAIN, FEATURE, DATE)) SPLIT ON ('EU','NA')

DATA GENERATOR
STRING :: Values: NA|CS|EU
STRING :: Values: Salesforce.com|Apple.com|Google.com|Yelp.com
STRING :: Values: Login|Report|Dashboard|Sales|UI
DATE :: Values: SEQUENTIAL
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 2000 Values: RANDOM
INTEGER :: Max. Value: 10000 Values: RANDOM

select count(*) from AGGREGATION_TABLE
1s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.41s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from AGGREGATION_TABLE group by HOST
1.1s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.49s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, count(1) from AGGREGATION_TABLE group by CORE
0.98s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.53s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, host, sum(ACTIVE_VISITOR) from AGGREGATION_TABLE group by core, host
2.36s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.41s 4.X-HBASE-0.98-0.98.17-HADOOP2

select AVG(DB), MIN(CORE), MAX(ACTIVE_VISITOR) from AGGREGATION_TABLE where host='CS' and DOMAIN<>'Google.com'
1.05s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.96s 4.X-HBASE-0.98-0.98.17-HADOOP2

select round(date, 'day',50), sum(core)+2 from AGGREGATION_TABLE where (CORE<10 or CORE>90) group by round(date, 'day',50), host
1.19s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.3s 4.X-HBASE-0.98-0.98.17-HADOOP2

select trunc(date, 'hour', 1000), count(DB), count(ACTIVE_VISITOR) from AGGREGATION_TABLE where feature='Login' and (core<50 or db<20 or active_visitor > 200) group by trunc(date, 'hour', 1000)
2.24s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.34s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(active_visitor) count from AGGREGATION_TABLE
2.08s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.08s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(distinct active_visitor) count from AGGREGATION_TABLE
2.78s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.84s 4.X-HBASE-0.98-0.98.17-HADOOP2

select percentile_cont(0.9) within group (order by active_visitor asc) from AGGREGATION_TABLE
2.75s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.87s 4.X-HBASE-0.98-0.98.17-HADOOP2

select percentile_cont(0.1) within group (order by active_visitor desc) from AGGREGATION_TABLE
2.75s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.85s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core), count(db), count(active_visitor) from AGGREGATION_TABLE
2.43s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.5s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(distinct core), count(distinct db), count(distinct active_visitor) from AGGREGATION_TABLE
3.51s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.68s 4.X-HBASE-0.98-0.98.17-HADOOP2

select percentile_cont(0.5) within group (order by core asc), percentile_cont(0.5) within group (order by db asc), percentile_cont(0.5) within group (order by active_visitor asc) from AGGREGATION_TABLE
3.53s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.74s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, count(db) count from AGGREGATION_TABLE group by CORE
0.97s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.85s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, count(distinct db) count from AGGREGATION_TABLE group by CORE
2.59s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.55s 4.X-HBASE-0.98-0.98.17-HADOOP2

select core, percentile_cont(0.9999) within group (order by db asc) from AGGREGATION_TABLE group by CORE
2.67s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.67s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core IN (1,25,50,75,100)
1.04s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.13s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core < 10 and db < 200
1.19s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.36s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core < 10 and active_visitor < 1000
2.03s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.06s 4.X-HBASE-0.98-0.98.17-HADOOP2

select * from AGGREGATION_TABLE LIMIT 100
0.02s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.02s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from AGGREGATION_TABLE where core > 90
1.09s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.12s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA AGGREGATION_TABLE [2000000 ROWS]
152.17s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
140.49s 4.X-HBASE-0.98-0.98.17-HADOOP2


IN_LIKE_TABLE


Phoenix 1.2+ uses skip scan that gives significant performance improvement for IN/LIKE queries.

ROWS 2000000 [2M]

DDL CREATE TABLE IF NOT EXISTS $TABLE (K1 CHAR(1) NOT NULL, K2 VARCHAR NOT NULL, CF.A INTEGER, CF.B VARCHAR, CF.C VARCHAR CONSTRAINT PK PRIMARY KEY (K1,K2)) SPLIT ON ('B','C','D','E','F','G','H','I')

DATA GENERATOR
STRING :: Values: A|B|C|D|E|F|G|H|I
STRING :: Length: 10 Values: SEQUENTIAL
INTEGER :: Max. Value: 1000000 Values: RANDOM
STRING :: Length: 50 Values: RANDOM
STRING :: Length: 50 Values: RANDOM

select count(1) from IN_LIKE_TABLE where K1 in ('Z', 'D', 'F', 'X')
0.25s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.37s 4.X-HBASE-0.98-0.98.17-HADOOP2

select * from IN_LIKE_TABLE where K1 in ('A','B') and K2 in ('xxx1000000','xxx2000000','xxx2000001')
0.01s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0s 4.X-HBASE-0.98-0.98.17-HADOOP2

select * from IN_LIKE_TABLE where K1 in ('A', 'C', 'E', 'F', 'H') and K2 like 'xxxxxx1%'
0.02s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.01s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from IN_LIKE_TABLE where K1 in ('X','A','B','C','D','E','F','G','H','I','Z') and K2 like '%1%'
1.75s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.3s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from IN_LIKE_TABLE where (K1 in ('A','E','F','G','H','I','Z') or K1 like 'E%') and (K2 like '%xx1%' or K2 like '%xxxxxx2%')
1.21s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.67s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA IN_LIKE_TABLE [2000000 ROWS]
106.82s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
104.25s 4.X-HBASE-0.98-0.98.17-HADOOP2


INDEXED_TABLE


Indexing test. Phoenix 2.0+ feature. See explain plan by hovering over version label to see which index table is used.

ROWS 2000000 [2M]

DDL CREATE TABLE $TABLE (HOST CHAR(2) NOT NULL,DOMAIN VARCHAR NOT NULL,FEATURE VARCHAR NOT NULL,DATE DATE NOT NULL,USAGE.CORE BIGINT,USAGE.DB BIGINT,STATS.ACTIVE_VISITOR INTEGER CONSTRAINT PK PRIMARY KEY (HOST, DOMAIN, FEATURE, DATE)) IMMUTABLE_ROWS=true,MAX_FILESIZE=30485760;CREATE INDEX ids1 ON $TABLE (CORE) SALT_BUCKETS=16;CREATE INDEX ids2 ON $TABLE (DB) SALT_BUCKETS=16;CREATE INDEX ids3 ON $TABLE (DB,ACTIVE_VISITOR) SALT_BUCKETS=16;CREATE INDEX ids4 ON $TABLE (CORE,DB,ACTIVE_VISITOR) SALT_BUCKETS=16;

DATA GENERATOR
STRING :: Values: NA|CS|EU
STRING :: Values: Salesforce.com|Apple.com|Google.com|Yelp.com
STRING :: Values: Login|Report|Dashboard|Sales|UI
DATE :: Values: SEQUENTIAL
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 2000 Values: RANDOM
INTEGER :: Max. Value: 10000 Values: RANDOM

select count(*) from INDEXED_TABLE
0.52s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.52s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(*) from INDEXED_TABLE
0.49s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.5s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(*) from INDEXED_TABLE
0.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.85s 4.X-HBASE-0.98-0.98.17-HADOOP2

select round(date, 'day',50), sum(core)+2 from INDEXED_TABLE where (CORE<10 or CORE>90) group by round(date, 'day',50), host
0.27s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.26s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ round(date, 'day',50), sum(core)+2 from INDEXED_TABLE where (CORE<10 or CORE>90) group by round(date, 'day',50), host
0.95s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.04s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core IN (1,100)
0.04s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.05s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(core) from INDEXED_TABLE where core IN (1,100)
0.03s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.04s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core IN (1,100)
0.7s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.81s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core IN (1,25,50,75,100)
0.07s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.07s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(core) from INDEXED_TABLE where core IN (1,25,50,75,100)
0.06s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.06s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core IN (1,25,50,75,100)
0.68s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.82s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core < 10 and db < 200
0.12s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.13s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+ INDEX(INDEXED_TABLE ids1 ids2 ids3 ids4) */ count(core) from INDEXED_TABLE where core < 10 and db < 200
11.09s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
11.1s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core < 10 and db < 200
0.83s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.97s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core < 100 and db < 2000
1.18s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.54s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core < 100 and db < 2000
1.16s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.39s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core > 90
0.07s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.08s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core > 90
0.73s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.83s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(core) from INDEXED_TABLE where core > 0
0.59s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.62s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ count(core) from INDEXED_TABLE where core > 0
0.85s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.97s 4.X-HBASE-0.98-0.98.17-HADOOP2

select db from INDEXED_TABLE where db >5 and db <7 order by db
0.04s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.03s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ db from INDEXED_TABLE where db >5 and db <7 order by db
1.32s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.02s 4.X-HBASE-0.98-0.98.17-HADOOP2

select db from INDEXED_TABLE order by db limit 10
0.05s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.02s 4.X-HBASE-0.98-0.98.17-HADOOP2

select /*+NO_INDEX*/ db from INDEXED_TABLE order by db limit 10
1.36s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.35s 4.X-HBASE-0.98-0.98.17-HADOOP2

CREATE INDEX idx5 ON INDEXED_TABLE (CORE) INCLUDE (DB,ACTIVE_VISITOR)
37.51s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
26.24s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA INDEXED_TABLE [2000000 ROWS]
383.92s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
367.53s 4.X-HBASE-0.98-0.98.17-HADOOP2


TABLE_6CF


Data is split in 6 column families. Phoenix 1.2+ uses Essential Column Family filter feature which fetches only required columns when filtering rows.

ROWS 2000000 [2M]

DDL CREATE TABLE IF NOT EXISTS $TABLE (K1 CHAR(1) NOT NULL, K2 VARCHAR NOT NULL, CF1.A INTEGER, CF2.B INTEGER, CF3.C INTEGER, CF4.D INTEGER, CF5.E INTEGER, CF6.F INTEGER CONSTRAINT PK PRIMARY KEY (K1,K2)) SPLIT ON ('B','C','D');CREATE TABLE IF NOT EXISTS $TABLE_UPSERTSELECT (K1 CHAR(1) NOT NULL, K2 VARCHAR NOT NULL, CF1.A INTEGER, CF2.B INTEGER, CF3.C INTEGER, CF4.D INTEGER, CF5.E INTEGER, CF6.F INTEGER CONSTRAINT PK PRIMARY KEY (K1,K2)) SPLIT ON ('B','C','D')

DATA GENERATOR
STRING :: Values: A|B|C|D
STRING :: Length: 10 Values: SEQUENTIAL
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 1000 Values: RANDOM
INTEGER :: Max. Value: 2000 Values: RANDOM
INTEGER :: Max. Value: 5000 Values: RANDOM
INTEGER :: Max. Value: 10000 Values: RANDOM
INTEGER :: Max. Value: 15000 Values: RANDOM

select count(1) from TABLE_6CF where K1 in ('A', 'B', 'C', 'D')
0.98s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.04s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from TABLE_6CF where K1 in ('A', 'C')
0.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.62s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_6CF where f < 1000
0.92s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.64s 4.X-HBASE-0.98-0.98.17-HADOOP2

select a,b,c,d,e,f from TABLE_6CF where B>1000 and B<2000 and f>1000 and f<2000
1.82s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.64s 4.X-HBASE-0.98-0.98.17-HADOOP2

select sum(a),sum(b),sum(c),sum(d) from TABLE_6CF group by K1
5.01s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
4.69s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_6CF group by A
0.72s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.93s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_6CF group by B
2.26s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.86s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_6CF where f < 10000
1.76s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.72s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA TABLE_6CF [2000000 ROWS]
145.43s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
142.44s 4.X-HBASE-0.98-0.98.17-HADOOP2


SALTED_TABLE


This is Phoenix 1.2+ feature. Data is split in 4 regions equal to number of region servers in test cluster by adding salted byte to each key.

ROWS 2000000 [2M]

DDL CREATE TABLE $TABLE (K1 CHAR(1) NOT NULL, K2 VARCHAR NOT NULL, CFA.A VARCHAR, CFA.B CHAR(100), CFA.C INTEGER, CFB.D INTEGER CONSTRAINT PK PRIMARY KEY (K1,K2)) SALT_BUCKETS=4

DATA GENERATOR
STRING :: Values: A|B|C|D
STRING :: Length: 10 Values: SEQUENTIAL
STRING :: Length: 100 Values: RANDOM
STRING :: Length: 100 Values: RANDOM
INTEGER :: Max. Value: 1000000 Values: RANDOM
INTEGER :: Max. Value: 1000000 Values: RANDOM

select count(1) from SALTED_TABLE where K1 in ('A', 'B', 'C', 'D')
1.37s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.89s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from SALTED_TABLE where D < 1000
0.68s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.77s 4.X-HBASE-0.98-0.98.17-HADOOP2

select avg(c),avg(d) from SALTED_TABLE group by K1
2.9s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.93s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA SALTED_TABLE [2000000 ROWS]
141.73s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
143.13s 4.X-HBASE-0.98-0.98.17-HADOOP2


TABLE_1CF


Same data as in 6 column family table but this one only has a single column family. Performance should be worse in most cases.

ROWS 2000000 [2M]

DDL CREATE TABLE IF NOT EXISTS $TABLE (K1 CHAR(1) NOT NULL, K2 VARCHAR NOT NULL, CF.A INTEGER, CF.B INTEGER, CF.C INTEGER, CF.D INTEGER, CF.E INTEGER, CF.F INTEGER CONSTRAINT PK PRIMARY KEY (K1,K2)) SPLIT ON ('B','C','D');CREATE TABLE IF NOT EXISTS $TABLE_UPSERTSELECT (K1 CHAR(1) NOT NULL, K2 VARCHAR NOT NULL, CF.A INTEGER, CF.B INTEGER, CF.C INTEGER, CF.D INTEGER, CF.E INTEGER, CF.F INTEGER CONSTRAINT PK PRIMARY KEY (K1,K2)) SPLIT ON ('B','C','D')

DATA GENERATOR
STRING :: Values: A|B|C|D
STRING :: Length: 10 Values: SEQUENTIAL
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 1000 Values: RANDOM
INTEGER :: Max. Value: 2000 Values: RANDOM
INTEGER :: Max. Value: 5000 Values: RANDOM
INTEGER :: Max. Value: 10000 Values: RANDOM
INTEGER :: Max. Value: 15000 Values: RANDOM

select count(1) from TABLE_1CF where K1 in ('A', 'B', 'C', 'D')
1.51s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.39s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(1) from TABLE_1CF where K1 in ('A', 'C')
1.1s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.74s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF where f < 1000
1.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.32s 4.X-HBASE-0.98-0.98.17-HADOOP2

select a,b,c,d,e,f from TABLE_1CF where B>1000 and B<2000 and f>1000 and f<2000
2.11s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.51s 4.X-HBASE-0.98-0.98.17-HADOOP2

select sum(a),sum(b),sum(c),sum(d) from TABLE_1CF group by K1
1.69s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.92s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF group by A
1.57s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.31s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF group by B
2.99s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.06s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from TABLE_1CF where f < 10000
1.61s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
1.4s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA TABLE_1CF [2000000 ROWS]
126.44s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
128.48s 4.X-HBASE-0.98-0.98.17-HADOOP2


WIDE_PK




ROWS 2000000 [2M]

DDL CREATE TABLE IF NOT EXISTS $TABLE (mypk CHAR(500) NOT NULL PRIMARY KEY,CF.column1 INTEGER,CF.column2 INTEGER,CF.column3 INTEGER,CF.column4 INTEGER,CF.column5 INTEGER) SALT_BUCKETS=4;

DATA GENERATOR
STRING :: Length: 500 Values: SEQUENTIAL
INTEGER :: Values: SEQUENTIAL
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 100 Values: RANDOM
INTEGER :: Max. Value: 100 Values: RANDOM

select count(*) from WIDE_PK
3.13s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.44s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from WIDE_PK where column2<10
2.53s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
2.5s 4.X-HBASE-0.98-0.98.17-HADOOP2

select * from WIDE_PK order by mypk DESC limit 10
0.12s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
0.43s 4.X-HBASE-0.98-0.98.17-HADOOP2

select count(*) from WIDE_PK where column2<50 and column3<50 and column4<50 and column5<50
2.83s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
3.27s 4.X-HBASE-0.98-0.98.17-HADOOP2

LOAD DATA WIDE_PK [2000000 ROWS]
284.28s V4.6.0-HBASE-0.98-0.98.17-HADOOP2
309.02s 4.X-HBASE-0.98-0.98.17-HADOOP2



Note: Hover over version label to see Phoenix Query Explain Plan.