[Input Speed]
Increase TPS as the number of input processes increased and the TPS is 1,090,000 for the input of 5 processes.
[Query Speed]
Measuring query speed with the conditions of time period among 7,000,000,000 stored data.
2. Overview of Performance Index
Criteria for Performance Evaluation of InfiniFlux
Data Storage Speed Query Speed
• Evaluating the speed based on the
number of data storage options
• Evaluating the speed based on the
amount of data input process
• Evaluating the speed of data count
• Evaluating the speed based on the
number of search conditions
•Evaluating the speed of operation f
unctions
2
3. Environments for Evaluation
• Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz (4-core)
• 32GB Memory
• 4 SAS Disks (7200 rpm)
Hardware
• Measuring speed based on the number of input data: 100,000,000, 500,000,000, 1,000,000,000,
3,000,000,000, 5,000,000,000, and 10,000,000,000 data records
• Measuring speed based on the amount of input process: 1~5 processes
Input Speed
• Run queries against 100,000,000 out of 7,000,000,000 data records
• Measuring speed based on the number of conditions
• Measuring speed over calculating the average value of data
Query Speed
3
4. Input Speed
Around 600,000 TPS for the input from 100,000,000 up to
10,000,000,000 data records
Number of records Time (sec) TPS
10,000,000 17 588235
50,000,000 86 581395
100,000,000 172 581395
300,000,000 503 596421
500,000,000 854 585480
700,000,000 1175 595744
1,000,000,000 1678 595947
588,235
581,395 581,395
596,421
585,480
595,744
595,947
570,000
575,000
580,000
585,000
590,000
595,000
600,000
100M 500M 1B 3B 5B 7B 10B
Input Speed
TPS
4
5. Input Speed
Increase TPS as the number of input processes increased and the TPS is
1,090,000 for the input of 5 processes
Number of process TPS
1 588235
2 754985
3 825825
4 1039136
5 1091993
588235
754985
825825
1039136
1091993
0
200000
400000
600000
800000
1000000
1200000
1 2 3 4 5
Input Speed
TPS
5
6. Query Speed
Measuring query speed with the conditions of time period among
7,000,000,000 stored data
Classification Query Statement
Result
(Number of
records)
Time (sec)
Full count with one condition
SELECT COUNT(*) FROM SENSOR_DATA_1 WHERE TYPE = 1
DURATION FROM to_date ('2015-09-03 19:05:00')
to to_date ('2015-09-03 19:10:00');
31,520,167 1.539
Full count with two conditions
SELECT COUNT(*) FROM SENSOR_DATA_1 WHERE
TYPE = 1 and ID = 10
DURATION FROM to_date ('2015-09-03 19:05:00')
to to_date ('2015-09-03 19:10:00');
315,202 0.663
6
7. Query Speed
Measuring query speed with the conditions of time period among
7,000,000,000 stored data
Classification Query Statement
Result (Number
of records)
Time (sec)
Count a specified
column with two
conditions
SELECT ID, COUNT(ID) FROM SENSOR_DATA_1 WHERE
TYPE = 1 and ID between 10 and 15
GROUP BY ID ORDER BY ID
DURATION FROM to_date ('2015-09-03 19:05:00')
to to_date ('2015-09-03 19:10:00');
ID COUNT(ID)
-------------------------
10 315202
11 315202
12 315202
13 315201
14 315202
15 315202
1.258
Calculate average values
with two conditions
SELECT TYPE,ID,AVG(VALUE) FROM SENSOR_DATA_1 WHERE TYPE
= 1 and ID = 10 group by TYPE,ID
DURATION FROM to_date ('2015-09-03 19:05:00')
to to_date ('2015-09-03 19:10:00');
TYPE ID AVG(VALUE)
-------------------------
1 10 22.4942
2.04
7
8. The World's Fastest
Time Series DBMS
for IoT and Big Data
www.infiniflux.com
info@infiniflux.com
InfiniFlux