PERFORMANCE
Just try it! Machbase has amazing data processing performance and ranked #1 in TPCx-IoT. Machbase DBMS, which is 3 to 17 times faster than others, take the best score with AMD in the TPCx-IoT evaluation.
Performance in AWS server
The Fastest Time Series DBMS for Indutrial IoT Data Machbase offers outstanding performance and dramatic powerful results.
Check out the amazing performance of Machbase Edge edition and Fog edition. Even if it is not a high-performance server, 1 million sensor data can be input per second and various statistical values can be inquired under second at the same time as data is stored. The outstanding performance of Machbase is possible due to the unique and innovative technology for real-time big data processing.
Inputs: Over 1 Million EPS! Queries: Less Than a 1 Second!
Input performance
Query performance
- Q1 : Elapsed time to retrieve a specific tag for 1 day
- Q2 : Average values per 60 seconds of a specific tag for 1 hour
- Q3 : Average values per 15 minutes of a specific tag for 1 day
Test environment
| Item | t3.xlarge instance | c5d.2xlarge instance |
| CPU | AIntel Skylake P-8175 @2.5Ghz 4 vCPUs | Intel Xeon Platinum 8124M @3Ghz 8 vCPUs |
| Memory | 16GB | 16GB |
| Disk | 180GB NVMe SSD | 200GB NVMe SSD |
| OS | Amazon Linux (Kernel 4.14.88-88.76.amzn2.x86_64) | Amazon Linux (Kernel 4.14.88-88.76.amzn2.x86_64) |
| NIC | 1Gbps | 1Gbps |
| Machbase | Machbase Edge 6.1 Release | Machbase Fog 6.1 Release |
Test results (input performance)
| Item | t3.xlarge instance | c5d.2xlarge instance |
| Total count of data | 5 billion (5,000,000,000) | 5 billion (5,000,000,000) |
| Data loading time (sec) | 12,391 seconds | 4,705 seconds |
| Inputs per second (Event / Sec) | 403,509 events /sec | 1,062,560 events /sec |
| CPU usage during data entry | 118% ~ 367% / (total 400%) | 275% ~ 616% / (total 800%) |
| Memory usage during data entry | RRES 7.0 ~ 7.6 GB / (total16GB) | RES 7.0 ~ 7.5 GB / (total16GB) |
| Disk Usage (GB) | 162 GB / (180GB total) | 162 GB / (200GB total) |
Test results (query performance)
Edge Edition
| Item | SQL QUERY | Results | Time (seconds) |
| total count of data | SELECT COUNT(*) FROM TAG; | 5000000000 | 0.000 |
| min and max of time range for all data | SELECT MIN(TIME), MAX(TIME) FROM TAG; | 2018-01-01 00:00:00, 2018-01-02 03:46:39 | 0.003 |
| min and max of time range for a specific tag | SELECT MIN(TIME), MAX(TIME) FROM TAG WHERE NAME = ‘EQ0^TAG567’; | 2018-01-01 00:00:00, 2018-01-02 03:46:39 | 0.085 |
| retrieve 1 tag for 1 day | SELECT COUNT(*) FROM(SELECT * FROM TAG WHERE NAME = "EQ0^TAG287" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59")); | 431,996 | 0.061 |
| retrieve 15 tags for 1 day | SELECT COUNT(*) FROM (SELECT * FROM TAG WHERE NAME IN (("EQ0^TAG17", "EQ0^TAG27", "EQ0^TAG37", "EQ0^TAG47", "EQ0^TAG57", "EQ0^TAG127", "EQ0^TAG137", "EQ0^TAG287", "EQ0^TAG398", "EQ0^TAG467", "EQ0^TAG455", "EQ0^TAG512", "EQ0^TAG677", "EQ0^TAG834", "EQ0^TAG934") AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59")); | 6,479,940 | 0.827 |
| average values per second of 1 tag for 8 hours | SELECT TIME ROLLUP 1 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG187" AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 19:59:59") GROUP BY TIME ORDER BY TIME; | 28,800 | 0.179 |
| sum values per minute of 1 tag for 1 hour | SELECT TIME ROLLUP 1 MIN AS TIME, SUM(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG678" AND TIME BETWEEN TO_DATE("2018-01-01 20:00:00") AND TO_DATE("2018-01-01 20:59:59") GROUP BY TIME ORDER BY TIME; | 60 | 0.001 |
| maximum values per hour of 1 tag for 1 day | SELECT TIME ROLLUP 1 HOUR AS TIME, MAX(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG24" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 24 | 0.001 |
| average values per 15 seconds of 1 tag for 1 hour | SELECT TIME ROLLUP 15 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG587" AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 12:59:59") GROUP BY TIME ORDER BY TIME; | 240 | 0.013 |
| average values per 30 seconds of 1 tag for 1 hour | SELECT TIME ROLLUP 30 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG287 AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 12:59:59") GROUP BY TIME ORDER BY TIME; | 120 | 0.010 |
| average values per 60 seconds of 1 tag for 1 hour | SELECT TIME ROLLUP 60 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG987" AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 12:59:59") GROUP BY TIME ORDER BY TIME; | 60 | 0.012 |
| average values per 5 minutes of 1 tag for 1 day | SELECT TIME ROLLUP 5 MIN AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG387" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 288 | 0.251 |
| average values per 10 minutes of 1 tag for 1 day | SELECT TIME ROLLUP 10 MIN AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG687" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 144 | 0.249 |
| average values per 15 minutes of 1 tag for 1 day | SELECT TIME ROLLUP 15 MIN AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG487" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 96 | 0.244 |
Fog Edition
| Item | SQL QUERY | Results | Time (seconds) |
| total count of data | SELECT COUNT(*) FROM TAG; | 5000000000 | 0.000 |
| min and max of time range for all data | SELECT MIN(TIME), MAX(TIME) FROM TAG; | 2018-01-01 00:00:00, 2018-01-02 03:46:39 | 0.003 |
| min and max of time range for a specific tag | SELECT MIN(TIME), MAX(TIME) FROM TAG WHERE NAME = ‘EQ0^TAG567’; | 2018-01-01 00:00:00, 2018-01-02 03:46:39 | 0.065 |
| retrieve 1 tag for 1 day | SELECT COUNT(*) FROM(SELECT * FROM TAG WHERE NAME = "EQ0^TAG287" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59")); | 431,996 | 0.053 |
| retrieve 15 tags for 1 day | SELECT COUNT(*) FROM (SELECT * FROM TAG WHERE NAME IN (("EQ0^TAG17", "EQ0^TAG27", "EQ0^TAG37", "EQ0^TAG47", "EQ0^TAG57", "EQ0^TAG127", "EQ0^TAG137", "EQ0^TAG287", "EQ0^TAG398", "EQ0^TAG467", "EQ0^TAG455", "EQ0^TAG512", "EQ0^TAG677", "EQ0^TAG834", "EQ0^TAG934") AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59")); | 6,479,940 | 0.723 |
| average values per second of 1 tag for 8 hours | SELECT TIME ROLLUP 1 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG187" AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 19:59:59") GROUP BY TIME ORDER BY TIME; | 28,800 | 0.121 |
| sum values per minute of 1 tag for 1 hour | SELECT TIME ROLLUP 1 MIN AS TIME, SUM(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG678" AND TIME BETWEEN TO_DATE("2018-01-01 20:00:00") AND TO_DATE("2018-01-01 20:59:59") GROUP BY TIME ORDER BY TIME; | 60 | 0.001 |
| maximum values per hour of 1 tag for 1 day | SELECT TIME ROLLUP 1 HOUR AS TIME, MAX(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG24" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 24 | 0.000 |
| average values per 15 seconds of 1 tag for 1 hour | SELECT TIME ROLLUP 15 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG587" AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 12:59:59") GROUP BY TIME ORDER BY TIME; | 240 | 0.009 |
| average values per 30 seconds of 1 tag for 1 hour | SELECT TIME ROLLUP 30 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG287 AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 12:59:59") GROUP BY TIME ORDER BY TIME; | 120 | 0.008 |
| average values per 60 seconds of 1 tag for 1 hour | SELECT TIME ROLLUP 60 SEC AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG987" AND TIME BETWEEN TO_DATE("2018-01-01 12:00:00") AND TO_DATE("2018-01-01 12:59:59") GROUP BY TIME ORDER BY TIME; | 60 | 0.008 |
| average values per 5 minutes of 1 tag for 1 day | SELECT TIME ROLLUP 5 MIN AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG387" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 288 | 0.194 |
| average values per 10 minutes of 1 tag for 1 day | SELECT TIME ROLLUP 10 MIN AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG687" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 144 | 0.191 |
| average values per 15 minutes of 1 tag for 1 day | SELECT TIME ROLLUP 15 MIN AS TIME, AVG(VALUE) AS VALUE FROM TAG WHERE NAME = "EQ0^TAG487" AND TIME BETWEEN TO_DATE("2018-01-01 00:00:00") AND TO_DATE("2018-01-01 23:59:59") GROUP BY TIME ORDER BY TIME; | 96 | 0.200 |
