Introduction
This is a report for the insert benchmark with 20M docs and 4 client(s). It is generated by scripts (bash, awk, sed) and Tufte might not be impressed. An overview of the insert benchmark is here and a short update is here. Below, by DBMS, I mean DBMS+version.config. An example is my8020.c10b40 where my means MySQL, 8020 is version 8.0.20 and c10b40 is the name for the configuration file.
The test server has 8 AMD cores, 16G RAM and an NVMe SSD. It is described here as the Beelink. The benchmark was run with 4 clients and there were 1 or 2 connections per client (1 for queries, 1 for inserts). It uses 4 tables, each client has a separate table. It loads 20M rows without secondary indexes, creates secondary indexes, loads another 20M rows then does 3 read+write tests for one hour each that do queries as fast as possible with 100, 500 and then 1000 writes/second/client concurrent with the queries. The database is cached by OS but not by Postgres. Clients and the DBMS share one server. The per-database configs are in the per-database subdirectories here.
The tested DBMS are:
- pg152_o3_native_lto.cx7_1g -Postgres 15.2, the cx7_1g config and the build used: -O3 -march=native -mtune=native
- pg16prebeta - Postgres 16 from the main branch at git sha 1ab763fc2, the cx7_1g config and the build used: -O3 -march=native -mtune=native
Contents
- Summary
- l.i0: load without secondary indexes
- l.x: create secondary indexes
- l.i1: continue load after secondary indexes created
- q100.1: range queries with 100 insert/s per client
- q500.1: range queries with 500 insert/s per client
- q1000.1: range queries with 1000 insert/s per client
Summary
The numbers are inserts/s for l.i0 and l.i1, indexed docs (or rows) /s for l.x and queries/s for q*.2. The values are the average rate over the entire test for inserts (IPS) and queries (QPS). The range of values for IPS and QPS is split into 3 parts: bottom 25%, middle 50%, top 25%. Values in the bottom 25% have a red background, values in the top 25% have a green background and values in the middle have no color. A gray background is used for values that can be ignored because the DBMS did not sustain the target insert rate. Red backgrounds are not used when the minimum value is within 80% of the max value.
| dbms | l.i0 | l.x | l.i1 | q100.1 | q500.1 | q1000.1 |
|---|---|---|---|---|---|---|
| 20m.pg152_o3_native_lto.cx7_1g | 259740 | 254430 | 66225 | 46240 | 44391 | 43172 |
| 20m.pg16prebeta.cx7_1g | 246914 | 248148 | 67340 | 45996 | 44081 | 42684 |
This table has relative throughput, throughput for the DBMS relative to the DBMS in the first line, using the absolute throughput from the previous table.
| dbms | l.i0 | l.x | l.i1 | q100.1 | q500.1 | q1000.1 |
|---|---|---|---|---|---|---|
| 20m.pg152_o3_native_lto.cx7_1g | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 20m.pg16prebeta.cx7_1g | 0.95 | 0.98 | 1.02 | 0.99 | 0.99 | 0.99 |
This lists the average rate of inserts/s for the tests that do inserts concurrent with queries. For such tests the query rate is listed in the table above. The read+write tests are setup so that the insert rate should match the target rate every second. Cells that are not at least 95% of the target have a red background to indicate a failure to satisfy the target.
| dbms | q100.1 | q500.1 | q1000.1 |
|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 399 | 1994 | 3989 |
| pg16prebeta.cx7_1g | 399 | 1994 | 3989 |
| target | 400 | 2000 | 4000 |
l.i0
l.i0: load without secondary indexes. Graphs for performance per 1-second interval are here.
Average throughput:
Insert response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For the max column values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 0.060 | 99.874 | 0.066 | 0.011 | |||||||
| pg16prebeta.cx7_1g | 99.948 | 0.052 | 0.014 |
Performance metrics for the DBMS listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 259740 0 0 0.0 405.5 131.4 0.000 0.000 0.002 0.518 28203 69.8 0.109 22 2.1 5.6 0.0 0.011 75746 1898 20m.pg152_o3_native_lto.cx7_1g 246914 0 0 0.0 393.6 124.5 0.000 0.000 0.002 0.516 27748 68.8 0.112 22 2.1 5.6 0.0 0.014 74419 31666 20m.pg16prebeta.cx7_1g
l.x
l.x: create secondary indexes.
Average throughput:
Performance metrics for the DBMS listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 254430 0 0 0.0 153.3 66.2 0.000 0.000 0.001 0.266 834 12.6 0.003 4 3.9 9.0 0.0 0.003 NA NA 20m.pg152_o3_native_lto.cx7_1g 248148 0 0 0.0 171.4 73.5 0.000 0.000 0.001 0.303 710 12.8 0.003 4 3.9 9.0 0.0 0.003 NA NA 20m.pg16prebeta.cx7_1g
l.i1
l.i1: continue load after secondary indexes created. Graphs for performance per 1-second interval are here.
Average throughput:
Insert response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For the max column values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 97.579 | 2.345 | 0.074 | 0.001 | 0.077 | ||||||
| pg16prebeta.cx7_1g | 98.147 | 1.778 | 0.074 | 0.001 | 0.078 |
Performance metrics for the DBMS listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 66225 0 179 1.3 43675.5 835.9 0.003 0.019 0.660 12.925 42728 62.3 0.645 75 8.6 25.3 0.0 0.077 16989 6043 20m.pg152_o3_native_lto.cx7_1g 67340 0 192 1.3 44741.5 858.3 0.003 0.020 0.664 13.051 40491 63.2 0.601 75 8.6 25.3 0.0 0.078 17381 7891 20m.pg16prebeta.cx7_1g
q100.1
q100.1: range queries with 100 insert/s per client. Graphs for performance per 1-second interval are here.
Average throughput:
Query response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For max values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 99.983 | 0.011 | 0.005 | nonzero | 0.009 | ||||||
| pg16prebeta.cx7_1g | 99.983 | 0.011 | 0.006 | nonzero | 0.008 |
Insert response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For max values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 99.528 | 0.472 | 0.012 | ||||||||
| pg16prebeta.cx7_1g | 99.688 | 0.312 | 0.011 |
Performance metrics for the DBMS listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 399 46240 2 0.0 1278.7 15.4 0.000 0.000 3.206 39.576 176341 50.6 3.814 88 8.8 22.4 0.0 0.009 11715 10580 20m.pg152_o3_native_lto.cx7_1g 399 45996 1 0.0 1276.2 15.4 0.000 0.000 3.200 39.506 175374 50.4 3.813 88 8.7 22.3 0.0 0.008 11592 10278 20m.pg16prebeta.cx7_1g
q500.1
q500.1: range queries with 500 insert/s per client. Graphs for performance per 1-second interval are here.
Average throughput:
Query response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For max values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 99.942 | 0.038 | 0.019 | 0.001 | 0.016 | ||||||
| pg16prebeta.cx7_1g | 99.944 | 0.035 | 0.019 | 0.002 | nonzero | 0.018 |
Insert response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For max values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 96.400 | 3.539 | 0.061 | 0.053 | |||||||
| pg16prebeta.cx7_1g | 91.306 | 8.629 | 0.065 | 0.027 |
Performance metrics for the DBMS listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 1994 44391 34 0.3 5763.2 55.3 0.001 0.006 2.890 28.376 167977 52.0 3.784 94 10.2 20.1 0.0 0.016 11172 9542 20m.pg152_o3_native_lto.cx7_1g 1994 44081 34 0.3 5759.6 55.2 0.001 0.006 2.888 28.363 166743 51.9 3.783 94 10.2 20.0 0.0 0.018 11044 9669 20m.pg16prebeta.cx7_1g
q1000.1
q1000.1: range queries with 1000 insert/s per client. Graphs for performance per 1-second interval are here.
Average throughput:
Query response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For max values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 99.903 | 0.069 | 0.026 | 0.002 | nonzero | 0.022 | |||||
| pg16prebeta.cx7_1g | 99.898 | 0.073 | 0.027 | 0.002 | nonzero | 0.018 |
Insert response time histogram: each cell has the percentage of responses that take <= the time in the header and max is the max response time in seconds. For max values in the top 25% of the range have a red background and in the bottom 25% of the range have a green background. The red background is not used when the min value is within 80% of the max value.
| dbms | 256us | 1ms | 4ms | 16ms | 64ms | 256ms | 1s | 4s | 16s | gt | max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| pg152_o3_native_lto.cx7_1g | 93.860 | 6.084 | 0.056 | 0.028 | |||||||
| pg16prebeta.cx7_1g | 94.599 | 5.337 | 0.065 | 0.028 |
Performance metrics for the DBMS listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 3989 43172 62 0.5 11390.0 100.4 0.001 0.013 2.855 25.777 160829 53.9 3.725 100 12.6 21.0 0.0 0.022 10772 9557 20m.pg152_o3_native_lto.cx7_1g 3989 42684 62 0.5 11377.0 100.3 0.001 0.013 2.852 25.744 159158 53.8 3.729 101 12.6 21.0 0.0 0.018 10804 9401 20m.pg16prebeta.cx7_1g
l.i0
l.i0: load without secondary indexes
Performance metrics for all DBMS, not just the ones listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 259740 0 0 0.0 405.5 131.4 0.000 0.000 0.002 0.518 28203 69.8 0.109 22 2.1 5.6 0.0 0.011 75746 1898 20m.pg152_o3_native_lto.cx7_1g 246914 0 0 0.0 393.6 124.5 0.000 0.000 0.002 0.516 27748 68.8 0.112 22 2.1 5.6 0.0 0.014 74419 31666 20m.pg16prebeta.cx7_1g
l.x
l.x: create secondary indexes
Performance metrics for all DBMS, not just the ones listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 254430 0 0 0.0 153.3 66.2 0.000 0.000 0.001 0.266 834 12.6 0.003 4 3.9 9.0 0.0 0.003 NA NA 20m.pg152_o3_native_lto.cx7_1g 248148 0 0 0.0 171.4 73.5 0.000 0.000 0.001 0.303 710 12.8 0.003 4 3.9 9.0 0.0 0.003 NA NA 20m.pg16prebeta.cx7_1g
l.i1
l.i1: continue load after secondary indexes created
Performance metrics for all DBMS, not just the ones listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 66225 0 179 1.3 43675.5 835.9 0.003 0.019 0.660 12.925 42728 62.3 0.645 75 8.6 25.3 0.0 0.077 16989 6043 20m.pg152_o3_native_lto.cx7_1g 67340 0 192 1.3 44741.5 858.3 0.003 0.020 0.664 13.051 40491 63.2 0.601 75 8.6 25.3 0.0 0.078 17381 7891 20m.pg16prebeta.cx7_1g
q100.1
q100.1: range queries with 100 insert/s per client
Performance metrics for all DBMS, not just the ones listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 399 46240 2 0.0 1278.7 15.4 0.000 0.000 3.206 39.576 176341 50.6 3.814 88 8.8 22.4 0.0 0.009 11715 10580 20m.pg152_o3_native_lto.cx7_1g 399 45996 1 0.0 1276.2 15.4 0.000 0.000 3.200 39.506 175374 50.4 3.813 88 8.7 22.3 0.0 0.008 11592 10278 20m.pg16prebeta.cx7_1g
q500.1
q500.1: range queries with 500 insert/s per client
Performance metrics for all DBMS, not just the ones listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 1994 44391 34 0.3 5763.2 55.3 0.001 0.006 2.890 28.376 167977 52.0 3.784 94 10.2 20.1 0.0 0.016 11172 9542 20m.pg152_o3_native_lto.cx7_1g 1994 44081 34 0.3 5759.6 55.2 0.001 0.006 2.888 28.363 166743 51.9 3.783 94 10.2 20.0 0.0 0.018 11044 9669 20m.pg16prebeta.cx7_1g
q1000.1
q1000.1: range queries with 1000 insert/s per client
Performance metrics for all DBMS, not just the ones listed above. Some are normalized by throughput, others are not. Legend for results is here.
ips qps rps rmbps wps wmbps rpq rkbpq wpi wkbpi csps cpups cspq cpupq dbgb1 dbgb2 rss maxop p50 p99 tag 3989 43172 62 0.5 11390.0 100.4 0.001 0.013 2.855 25.777 160829 53.9 3.725 100 12.6 21.0 0.0 0.022 10772 9557 20m.pg152_o3_native_lto.cx7_1g 3989 42684 62 0.5 11377.0 100.3 0.001 0.013 2.852 25.744 159158 53.8 3.729 101 12.6 21.0 0.0 0.018 10804 9401 20m.pg16prebeta.cx7_1g
l.i0
- l.i0: load without secondary indexes
- Legend for results is here.
- Each entry lists the percentage of responses that fit in that bucket (slower than max time for previous bucket, faster than min time for next bucket).
Insert response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 0.000 0.060 99.874 0.066 0.000 0.000 0.000 0.000 0.000 0.000 0.011 pg152_o3_native_lto.cx7_1g 0.000 0.000 99.948 0.052 0.000 0.000 0.000 0.000 0.000 0.000 0.014 pg16prebeta.cx7_1g
l.x
- l.x: create secondary indexes
- Legend for results is here.
- Each entry lists the percentage of responses that fit in that bucket (slower than max time for previous bucket, faster than min time for next bucket).
TODO - determine whether there is data for create index response time
l.i1
- l.i1: continue load after secondary indexes created
- Legend for results is here.
- Each entry lists the percentage of responses that fit in that bucket (slower than max time for previous bucket, faster than min time for next bucket).
Insert response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 0.000 0.000 97.579 2.345 0.074 0.001 0.000 0.000 0.000 0.000 0.077 pg152_o3_native_lto.cx7_1g 0.000 0.000 98.147 1.778 0.074 0.001 0.000 0.000 0.000 0.000 0.078 pg16prebeta.cx7_1g
q100.1
- q100.1: range queries with 100 insert/s per client
- Legend for results is here.
- Each entry lists the percentage of responses that fit in that bucket (slower than max time for previous bucket, faster than min time for next bucket).
Query response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 99.983 0.011 0.005 nonzero 0.000 0.000 0.000 0.000 0.000 0.000 0.009 pg152_o3_native_lto.cx7_1g 99.983 0.011 0.006 nonzero 0.000 0.000 0.000 0.000 0.000 0.000 0.008 pg16prebeta.cx7_1g
Insert response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 0.000 0.000 99.528 0.472 0.000 0.000 0.000 0.000 0.000 0.000 0.012 pg152_o3_native_lto.cx7_1g 0.000 0.000 99.688 0.312 0.000 0.000 0.000 0.000 0.000 0.000 0.011 pg16prebeta.cx7_1g
q500.1
- q500.1: range queries with 500 insert/s per client
- Legend for results is here.
- Each entry lists the percentage of responses that fit in that bucket (slower than max time for previous bucket, faster than min time for next bucket).
Query response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 99.942 0.038 0.019 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.016 pg152_o3_native_lto.cx7_1g 99.944 0.035 0.019 0.002 nonzero 0.000 0.000 0.000 0.000 0.000 0.018 pg16prebeta.cx7_1g
Insert response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 0.000 0.000 96.400 3.539 0.061 0.000 0.000 0.000 0.000 0.000 0.053 pg152_o3_native_lto.cx7_1g 0.000 0.000 91.306 8.629 0.065 0.000 0.000 0.000 0.000 0.000 0.027 pg16prebeta.cx7_1g
q1000.1
- q1000.1: range queries with 1000 insert/s per client
- Legend for results is here.
- Each entry lists the percentage of responses that fit in that bucket (slower than max time for previous bucket, faster than min time for next bucket).
Query response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 99.903 0.069 0.026 0.002 nonzero 0.000 0.000 0.000 0.000 0.000 0.022 pg152_o3_native_lto.cx7_1g 99.898 0.073 0.027 0.002 nonzero 0.000 0.000 0.000 0.000 0.000 0.018 pg16prebeta.cx7_1g
Insert response time histogram
256us 1ms 4ms 16ms 64ms 256ms 1s 4s 16s gt max tag 0.000 0.000 93.860 6.084 0.056 0.000 0.000 0.000 0.000 0.000 0.028 pg152_o3_native_lto.cx7_1g 0.000 0.000 94.599 5.337 0.065 0.000 0.000 0.000 0.000 0.000 0.028 pg16prebeta.cx7_1g