产品性能
最近更新时间: 2026-06-30 15:06:00
使用 Elasticsearch 官方提供的 rally 压测工具和 geonames 数据集（大小3.2G，共11523468个文档），对TCloudFinanceZone Elasticsearch Service 构建的两种规格的 ES 集群进行压测，性能结果如下：
2核8G 3个节点：
| Metric | Task | Value | Unit |
|---|---|---|---|
| Total indexing time | - | 18.5837 | min |
| Min indexing time per shard | - | 0.000533333 | min |
| Median indexing time per shard | - | 3.4134 | min |
| Max indexing time per shard | - | 4.21195 | min |
| Total refresh time | - | 2.40147 | min |
| Min refresh time per shard | - | 0.0009 | min |
| Median refresh time per shard | - | 0.45715 | min |
| Max refresh time per shard | - | 0.5655 | min |
| Total flush time | - | 0.328317 | min |
| Min flush time per shard | - | 0.0004 | min |
| Median flush time per shard | - | 0.0643583 | min |
| Max flush time per shard | - | 0.0972167 | min |
| Total Young Gen GC | - | 127.992 | s |
| Total Old Gen GC | - | 1.327 | s |
| Store size | - | 3.22145 | GB |
| Translog size | - | 2.80E-07 | GB |
| Heap used for segments | - | 10.8161 | MB |
| Heap used for doc values | - | 0.0831299 | MB |
| Heap used for terms | - | 9.62618 | MB |
| Heap used for norms | - | 0.0795288 | MB |
| Heap used for points | - | 0.217488 | MB |
| Heap used for stored fields | - | 0.809799 | MB |
| Segment count | - | 102 | - |
| Min Throughput | index-append | 41895.6 | docs/s |
| Median Throughput | index-append | 42562 | docs/s |
| Max Throughput | index-append | 43352 | docs/s |
| 50th percentile latency | index-append | 757.636 | ms |
| 90th percentile latency | index-append | 1388.11 | ms |
| 99th percentile latency | index-append | 2650.44 | ms |
| 99.9th percentile latency | index-append | 5270.13 | ms |
| 100th percentile latency | index-append | 6535.29 | ms |
| 50th percentile service time | index-append | 757.636 | ms |
| 90th percentile service time | index-append | 1388.11 | ms |
| 99th percentile service time | index-append | 2650.44 | ms |
| 99.9th percentile service time | index-append | 5270.13 | ms |
| 100th percentile service time | index-append | 6535.29 | ms |
| error rate | index-append | 0 | % |
| Min Throughput | index-stats | 99.99 | ops/s |
| Median Throughput | index-stats | 100.04 | ops/s |
| Max Throughput | index-stats | 100.06 | ops/s |
| 50th percentile latency | index-stats | 6.04131 | ms |
| 90th percentile latency | index-stats | 6.56561 | ms |
| 99th percentile latency | index-stats | 10.7941 | ms |
| 99.9th percentile latency | index-stats | 22.6768 | ms |
| 100th percentile latency | index-stats | 24.5623 | ms |
| 50th percentile service time | index-stats | 5.9341 | ms |
| 90th percentile service time | index-stats | 6.42644 | ms |
| 99th percentile service time | index-stats | 7.56809 | ms |
| 99.9th percentile service time | index-stats | 22.5948 | ms |
| 100th percentile service time | index-stats | 24.4467 | ms |
| error rate | index-stats | 0 | % |
| Min Throughput | node-stats | 99.93 | ops/s |
| Median Throughput | node-stats | 100.06 | ops/s |
| Max Throughput | node-stats | 100.33 | ops/s |
| 50th percentile latency | node-stats | 6.74088 | ms |
| 90th percentile latency | node-stats | 7.28822 | ms |
| 99th percentile latency | node-stats | 8.62256 | ms |
| 99.9th percentile latency | node-stats | 13.271 | ms |
| 100th percentile latency | node-stats | 13.9379 | ms |
| 50th percentile service time | node-stats | 6.64634 | ms |
| 90th percentile service time | node-stats | 7.18403 | ms |
| 99th percentile service time | node-stats | 8.34209 | ms |
| 99.9th percentile service time | node-stats | 13.1784 | ms |
| 100th percentile service time | node-stats | 13.8411 | ms |
| error rate | node-stats | 0 | % |
| Min Throughput | default | 46.85 | ops/s |
| Median Throughput | default | 47.58 | ops/s |
| Max Throughput | default | 47.7 | ops/s |
| 50th percentile latency | default | 1023.59 | ms |
| 90th percentile latency | default | 1890.67 | ms |
| 99th percentile latency | default | 2017.58 | ms |
| 99.9th percentile latency | default | 2026.21 | ms |
| 100th percentile latency | default | 2026.21 | ms |
| 50th percentile service time | default | 20.2853 | ms |
| 90th percentile service time | default | 24.3425 | ms |
| 99th percentile service time | default | 33.3526 | ms |
| 99.9th percentile service time | default | 54.702 | ms |
| 100th percentile service time | default | 74.2832 | ms |
| error rate | default | 0 | % |
| Min Throughput | term | 199.82 | ops/s |
| Median Throughput | term | 200.02 | ops/s |
| Max Throughput | term | 200.05 | ops/s |
| 50th percentile latency | term | 4.54844 | ms |
| 90th percentile latency | term | 9.97587 | ms |
| 99th percentile latency | term | 18.3546 | ms |
| 99.9th percentile latency | term | 21.2112 | ms |
| 100th percentile latency | term | 21.5267 | ms |
| 50th percentile service time | term | 4.39147 | ms |
| 90th percentile service time | term | 4.84527 | ms |
| 99th percentile service time | term | 6.62189 | ms |
| 99.9th percentile service time | term | 19.4932 | ms |
| 100th percentile service time | term | 21.4435 | ms |
| error rate | term | 0 | % |
| Min Throughput | phrase | 199.68 | ops/s |
| Median Throughput | phrase | 200.03 | ops/s |
| Max Throughput | phrase | 200.11 | ops/s |
| 50th percentile latency | phrase | 3.57488 | ms |
| 90th percentile latency | phrase | 4.62139 | ms |
| 99th percentile latency | phrase | 19.6935 | ms |
| 99.9th percentile latency | phrase | 24.9076 | ms |
| 100th percentile latency | phrase | 25.0486 | ms |
| 50th percentile service time | phrase | 3.4742 | ms |
| 90th percentile service time | phrase | 4.0265 | ms |
| 99th percentile service time | phrase | 7.57333 | ms |
| 99.9th percentile service time | phrase | 18.9011 | ms |
| 100th percentile service time | phrase | 23.8045 | ms |
| error rate | phrase | 0 | % |
| Min Throughput | country_agg_uncached | 4.99 | ops/s |
| Median Throughput | country_agg_uncached | 5 | ops/s |
| Max Throughput | country_agg_uncached | 5 | ops/s |
| 50th percentile latency | country_agg_uncached | 197.534 | ms |
| 90th percentile latency | country_agg_uncached | 217.842 | ms |
| 99th percentile latency | country_agg_uncached | 271.988 | ms |
| 100th percentile latency | country_agg_uncached | 275.963 | ms |
| 50th percentile service time | country_agg_uncached | 194.061 | ms |
| 90th percentile service time | country_agg_uncached | 209.086 | ms |
| 99th percentile service time | country_agg_uncached | 216.432 | ms |
| 100th percentile service time | country_agg_uncached | 217.275 | ms |
| error rate | country_agg_uncached | 0 | % |
| Min Throughput | country_agg_cached | 99.97 | ops/s |
| Median Throughput | country_agg_cached | 100.05 | ops/s |
| Max Throughput | country_agg_cached | 100.08 | ops/s |
| 50th percentile latency | country_agg_cached | 4.9212 | ms |
| 90th percentile latency | country_agg_cached | 5.44065 | ms |
| 99th percentile latency | country_agg_cached | 7.15509 | ms |
| 99.9th percentile latency | country_agg_cached | 16.9407 | ms |
| 100th percentile latency | country_agg_cached | 17.8111 | ms |
| 50th percentile service time | country_agg_cached | 4.81515 | ms |
| 90th percentile service time | country_agg_cached | 5.29377 | ms |
| 99th percentile service time | country_agg_cached | 6.38482 | ms |
| 99.9th percentile service time | country_agg_cached | 16.8318 | ms |
| 100th percentile service time | country_agg_cached | 17.7311 | ms |
| error rate | country_agg_cached | 0 | % |
| Min Throughput | scroll | 25.02 | pages/s |
| Median Throughput | scroll | 25.02 | pages/s |
| Max Throughput | scroll | 25.03 | pages/s |
| 50th percentile latency | scroll | 760.634 | ms |
| 90th percentile latency | scroll | 794.699 | ms |
| 99th percentile latency | scroll | 864.897 | ms |
| 100th percentile latency | scroll | 874.768 | ms |
| 50th percentile service time | scroll | 760.32 | ms |
| 90th percentile service time | scroll | 794.397 | ms |
| 99th percentile service time | scroll | 864.658 | ms |
| 100th percentile service time | scroll | 874.556 | ms |
| error rate | scroll | 0 | % |
| Min Throughput | expression | 2 | ops/s |
| Median Throughput | expression | 2 | ops/s |
| Max Throughput | expression | 2 | ops/s |
| 50th percentile latency | expression | 382.483 | ms |
| 90th percentile latency | expression | 414.775 | ms |
| 99th percentile latency | expression | 455.236 | ms |
| 100th percentile latency | expression | 473.181 | ms |
| 50th percentile service time | expression | 382.298 | ms |
| 90th percentile service time | expression | 414.577 | ms |
| 99th percentile service time | expression | 455.11 | ms |
| 100th percentile service time | expression | 472.998 | ms |
| error rate | expression | 0 | % |
| Min Throughput | painless_static | 1.5 | ops/s |
| Median Throughput | painless_static | 1.5 | ops/s |
| Max Throughput | painless_static | 1.5 | ops/s |
| 50th percentile latency | painless_static | 480.188 | ms |
| 90th percentile latency | painless_static | 505.003 | ms |
| 99th percentile latency | painless_static | 529.066 | ms |
| 100th percentile latency | painless_static | 547.199 | ms |
| 50th percentile service time | painless_static | 479.938 | ms |
| 90th percentile service time | painless_static | 504.731 | ms |
| 99th percentile service time | painless_static | 528.857 | ms |
| 100th percentile service time | painless_static | 546.954 | ms |
| error rate | painless_static | 0 | % |
| Min Throughput | painless_dynamic | 1.5 | ops/s |
| Median Throughput | painless_dynamic | 1.5 | ops/s |
| Max Throughput | painless_dynamic | 1.5 | ops/s |
| 50th percentile latency | painless_dynamic | 469.434 | ms |
| 90th percentile latency | painless_dynamic | 508.615 | ms |
| 99th percentile latency | painless_dynamic | 581.127 | ms |
| 100th percentile latency | painless_dynamic | 621.998 | ms |
| 50th percentile service time | painless_dynamic | 469.178 | ms |
| 90th percentile service time | painless_dynamic | 508.349 | ms |
| 99th percentile service time | painless_dynamic | 580.819 | ms |
| 100th percentile service time | painless_dynamic | 621.799 | ms |
| error rate | painless_dynamic | 0 | % |
| error rate | large_terms | 0 | % |
| Min Throughput | large_filtered_terms | 1.52 | ops/s |
| Median Throughput | large_filtered_terms | 1.52 | ops/s |
| Max Throughput | large_filtered_terms | 1.52 | ops/s |
| 50th percentile latency | large_filtered_terms | 39664.6 | ms |
| 90th percentile latency | large_filtered_terms | 46001.3 | ms |
| 99th percentile latency | large_filtered_terms | 47328 | ms |
| 100th percentile latency | large_filtered_terms | 47488.6 | ms |
| 50th percentile service time | large_filtered_terms | 651.731 | ms |
| 90th percentile service time | large_filtered_terms | 673.319 | ms |
| 99th percentile service time | large_filtered_terms | 715.941 | ms |
| 100th percentile service time | large_filtered_terms | 723.06 | ms |
| error rate | large_filtered_terms | 0 | % |
| Min Throughput | large_prohibited_terms | 1.55 | ops/s |
| Median Throughput | large_prohibited_terms | 1.56 | ops/s |
| Max Throughput | large_prohibited_terms | 1.57 | ops/s |
| 50th percentile latency | large_prohibited_terms | 35606.8 | ms |
| 90th percentile latency | large_prohibited_terms | 40847.6 | ms |
| 99th percentile latency | large_prohibited_terms | 42170.5 | ms |
| 100th percentile latency | large_prohibited_terms | 42329.2 | ms |
| 50th percentile service time | large_prohibited_terms | 648.82 | ms |
| 90th percentile service time | large_prohibited_terms | 672.114 | ms |
| 99th percentile service time | large_prohibited_terms | 722.666 | ms |
| 100th percentile service time | large_prohibited_terms | 733.307 | ms |
| error rate | large_prohibited_terms | 0 | % |
4核16G 3个节点：
| Metric | Task | Value | Unit |
|---|---|---|---|
| Total indexing time | - | 20.1957 | min |
| Min indexing time per shard | - | 0.000733333 | min |
| Median indexing time per shard | - | 3.77953 | min |
| Max indexing time per shard | - | 4.63752 | min |
| Total merge time | - | 1.57487 | min |
| Min merge time per shard | - | 0 | min |
| Median merge time per shard | - | 0.176658 | min |
| Max merge time per shard | - | 0.634067 | min |
| Total merge throttle time | - | 0.55105 | min |
| Min merge throttle time per shard | - | 0 | min |
| Median merge throttle time per shard | - | 0.065 | min |
| Max merge throttle time per shard | - | 0.217033 | min |
| Total refresh time | - | 1.41135 | min |
| Min refresh time per shard | - | 0.00106667 | min |
| Median refresh time per shard | - | 0.269958 | min |
| Max refresh time per shard | - | 0.345733 | min |
| Total flush time | - | 0.533133 | min |
| Min flush time per shard | - | 0.000566667 | min |
| Median flush time per shard | - | 0.115592 | min |
| Max flush time per shard | - | 0.136683 | min |
| Total Young Gen GC | - | 70.747 | s |
| Total Old Gen GC | - | 0.92 | s |
| Store size | - | 3.31581 | GB |
| Translog size | - | 2.80E-07 | GB |
| Heap used for segments | - | 11.0486 | MB |
| Heap used for doc values | - | 0.100529 | MB |
| Heap used for terms | - | 9.84413 | MB |
| Heap used for norms | - | 0.0755005 | MB |
| Heap used for points | - | 0.216421 | MB |
| Heap used for stored fields | - | 0.811981 | MB |
| Segment count | - | 97 | - |
| Min Throughput | index-append | 74421.1 | docs/s |
| Median Throughput | index-append | 75636.9 | docs/s |
| Max Throughput | index-append | 76877.4 | docs/s |
| 50th percentile latency | index-append | 377.922 | ms |
| 90th percentile latency | index-append | 663.055 | ms |
| 99th percentile latency | index-append | 3068.99 | ms |
| 100th percentile latency | index-append | 5554.97 | ms |
| 50th percentile service time | index-append | 377.922 | ms |
| 90th percentile service time | index-append | 663.055 | ms |
| 99th percentile service time | index-append | 3068.99 | ms |
| 100th percentile service time | index-append | 5554.97 | ms |
| error rate | index-append | 0 | % |
| Min Throughput | index-stats | 99.93 | ops/s |
| Median Throughput | index-stats | 100.04 | ops/s |
| Max Throughput | index-stats | 100.06 | ops/s |
| 50th percentile latency | index-stats | 6.62305 | ms |
| 90th percentile latency | index-stats | 7.35102 | ms |
| 99th percentile latency | index-stats | 18.0909 | ms |
| 99.9th percentile latency | index-stats | 24.5381 | ms |
| 100th percentile latency | index-stats | 24.7431 | ms |
| 50th percentile service time | index-stats | 6.50957 | ms |
| 90th percentile service time | index-stats | 7.18652 | ms |
| 99th percentile service time | index-stats | 9.38455 | ms |
| 99.9th percentile service time | index-stats | 24.4424 | ms |
| 100th percentile service time | index-stats | 24.6576 | ms |
| error rate | index-stats | 0 | % |
| Min Throughput | node-stats | 99.92 | ops/s |
| Median Throughput | node-stats | 100.04 | ops/s |
| Max Throughput | node-stats | 100.25 | ops/s |
| 50th percentile latency | node-stats | 7.15655 | ms |
| 90th percentile latency | node-stats | 7.96104 | ms |
| 99th percentile latency | node-stats | 10.2362 | ms |
| 99.9th percentile latency | node-stats | 25.7397 | ms |
| 100th percentile latency | node-stats | 29.1573 | ms |
| 50th percentile service time | node-stats | 7.04389 | ms |
| 90th percentile service time | node-stats | 7.84655 | ms |
| 99th percentile service time | node-stats | 9.13249 | ms |
| 99.9th percentile service time | node-stats | 10.7357 | ms |
| 100th percentile service time | node-stats | 29.072 | ms |
| error rate | node-stats | 0 | % |
| Min Throughput | default | 41.88 | ops/s |
| Median Throughput | default | 42.18 | ops/s |
| Max Throughput | default | 42.52 | ops/s |
| 50th percentile latency | default | 3789.31 | ms |
| 90th percentile latency | default | 5170.79 | ms |
| 99th percentile latency | default | 5582.03 | ms |
| 99.9th percentile latency | default | 5610.7 | ms |
| 100th percentile latency | default | 5618.05 | ms |
| 50th percentile service time | default | 23.1496 | ms |
| 90th percentile service time | default | 25.8865 | ms |
| 99th percentile service time | default | 33.249 | ms |
| 99.9th percentile service time | default | 45.493 | ms |
| 100th percentile service time | default | 62.4174 | ms |
| error rate | default | 0 | % |
| Min Throughput | term | 199.08 | ops/s |
| Median Throughput | term | 200 | ops/s |
| Max Throughput | term | 200.03 | ops/s |
| 50th percentile latency | term | 5.02391 | ms |
| 90th percentile latency | term | 21.18 | ms |
| 99th percentile latency | term | 35.2251 | ms |
| 99.9th percentile latency | term | 37.4827 | ms |
| 100th percentile latency | term | 37.6907 | ms |
| 50th percentile service time | term | 4.61812 | ms |
| 90th percentile service time | term | 5.10619 | ms |
| 99th percentile service time | term | 6.8135 | ms |
| 99.9th percentile service time | term | 22.1183 | ms |
| 100th percentile service time | term | 25.0033 | ms |
| error rate | term | 0 | % |
| Min Throughput | phrase | 199.61 | ops/s |
| Median Throughput | phrase | 200.04 | ops/s |
| Max Throughput | phrase | 200.8 | ops/s |
| 50th percentile latency | phrase | 3.86572 | ms |
| 90th percentile latency | phrase | 4.96583 | ms |
| 99th percentile latency | phrase | 22.5681 | ms |
| 99.9th percentile latency | phrase | 33.5684 | ms |
| 100th percentile latency | phrase | 28.2658 | ms |
| 50th percentile service time | phrase | 3.5689 | ms |
| 90th percentile service time | phrase | 4.2535 | ms |
| 99th percentile service time | phrase | 8.6957 | ms |
| 99.9th percentile service time | phrase | 24.5685 | ms |
| 100th percentile service time | phrase | 27.6584 | ms |
| error rate | phrase | 0 | % |
| Min Throughput | country_agg_uncached | 4.99 | ops/s |
| Median Throughput | country_agg_uncached | 5 | ops/s |
| Max Throughput | country_agg_uncached | 5 | ops/s |
| 50th percentile latency | country_agg_uncached | 182.291 | ms |
| 90th percentile latency | country_agg_uncached | 201.585 | ms |
| 99th percentile latency | country_agg_uncached | 257.343 | ms |
| 100th percentile latency | country_agg_uncached | 267.904 | ms |
| 50th percentile service time | country_agg_uncached | 181.161 | ms |
| 90th percentile service time | country_agg_uncached | 196.189 | ms |
| 99th percentile service time | country_agg_uncached | 216.762 | ms |
| 100th percentile service time | country_agg_uncached | 267.778 | ms |
| error rate | country_agg_uncached | 0 | % |
| Min Throughput | country_agg_cached | 99.95 | ops/s |
| Median Throughput | country_agg_cached | 100.05 | ops/s |
| Max Throughput | country_agg_cached | 100.07 | ops/s |
| 50th percentile latency | country_agg_cached | 5.57249 | ms |
| 90th percentile latency | country_agg_cached | 6.47982 | ms |
| 99th percentile latency | country_agg_cached | 9.33674 | ms |
| 99.9th percentile latency | country_agg_cached | 27.5319 | ms |
| 100th percentile latency | country_agg_cached | 32.0567 | ms |
| 50th percentile service time | country_agg_cached | 5.4601 | ms |
| 90th percentile service time | country_agg_cached | 6.25153 | ms |
| 99th percentile service time | country_agg_cached | 7.83564 | ms |
| 99.9th percentile service time | country_agg_cached | 13.6439 | ms |
| 100th percentile service time | country_agg_cached | 31.9487 | ms |
| error rate | country_agg_cached | 0 | % |
| Min Throughput | scroll | 25.01 | pages/s |
| Median Throughput | scroll | 25.03 | pages/s |