...
| Expand |
|---|
| title | page size vs response time |
|---|
|
Over time DAPI (1000 threads, 500 comments) slowed down over time. 
| Name | reqs | # | fails | Avg | Min | Max | Median | req/s | 95% |
|---|
| GET | GET_comment_list | 15793 | 0(0.00%) | 220 | 162 | 3116 | 200 | 0.2 | 340 | | GET | GET_thread | 955097 | 1(0.00%) | 214 | -8 | 4509 | 200 | 23.8 | 290 | | GET | GET_thread_list | 510921 | 0(0.00%) | 445 | 15 | 5595 | 420 | 11.6 | 810 | | PATCH | PATCH_comment | 15792 | 0(0.00%) | 337 | 187 | 3851 | 320 | 0.2 | 490 | | PATCH | PATCH_thread | 17333 | 164(0.94%) | 273 | 161 | 3019 | 250 | 0.5 | 400 |
When running a similar test with a more verbose output, we clearly saw that the page size had an effect on the response time. Although not shown here, larger content size was somewhat correlated to slower response times, but page_size had the better correlation. | Name | reqs | # | fails | Avg | Min | Max | Median | req/s | 95% |
|---|
| GET | page=1&page_size=100 | 238 | 0(0.00%) | 791 | 655 | 1990 | 750 | 0.4 | 960 | | GET | page=2&page_size=100 | 282 | 0(0.00%) | 762 | 659 | 1526 | 730 | 0.4 | 910 | | GET | page=1&page_size=75 | 240 | 0(0.00%) | 651 | 544 | 1038 | 620 | 0.2 | 830 | | GET | page=2&page_size=75 | 255 | 0(0.00%) | 677 | 573 | 1628 | 650 | 0 | 820 | | GET | page=7&page_size=100 | 260 | 0(0.00%) | 568 | 461 | 2637 | 520 | 0.4 | 710 | | GET | page=6&page_size=100 | 232 | 0(0.00%) | 559 | 466 | 843 | 520 | 0.1 | 710 | | GET | page=9&page_size=100 | 222 | 0(0.00%) | 568 | 460 | 1907 | 530 | 0.5 | 700 | | GET | page=3&page_size=100 | 262 | 0(0.00%) | 556 | 462 | 1807 | 520 | 0.2 | 700 | | GET | page=5&page_size=100 | 259 | 0(0.00%) | 556 | 463 | 1375 | 510 | 0.3 | 700 | | GET | page=4&page_size=100 | 276 | 0(0.00%) | 555 | 456 | 840 | 520 | 0.2 | 690 | | GET | page=8&page_size=100 | 236 | 0(0.00%) | 548 | 468 | 903 | 520 | 0.1 | 680 | | GET | page=3&page_size=75 | 251 | 0(0.00%) | 543 | 445 | 1348 | 510 | 0.3 | 670 | | GET | page=2&page_size=50 | 278 | 0(0.00%) | 507 | 415 | 1141 | 480 | 0.5 | 670 | | GET | page=1&page_size=50 | 251 | 0(0.00%) | 512 | 425 | 1043 | 490 | 0.1 | 660 | | GET | page=8&page_size=75 | 237 | 0(0.00%) | 481 | 397 | 1644 | 440 | 0.1 | 650 | | GET | page=3&page_size=50 | 247 | 0(0.00%) | 523 | 451 | 903 | 500 | 0.4 | 640 | | GET | page=6&page_size=75 | 238 | 0(0.00%) | 492 | 398 | 1634 | 450 | 0.3 | 640 | | GET | page=4&page_size=75 | 218 | 0(0.00%) | 480 | 395 | 886 | 440 | 0.2 | 630 | | GET | page=7&page_size=75 | 227 | 0(0.00%) | 474 | 391 | 795 | 440 | 0.5 | 620 | | GET | page=5&page_size=75 | 239 | 0(0.00%) | 466 | 388 | 1394 | 440 | 0.2 | 610 | | GET | page=4&page_size=50 | 237 | 0(0.00%) | 453 | 371 | 1010 | 420 | 0.2 | 610 | | GET | page=9&page_size=75 | 216 | 0(0.00%) | 467 | 397 | 996 | 430 | 0.4 | 600 | | GET | page=6&page_size=50 | 252 | 0(0.00%) | 391 | 321 | 1065 | 360 | 0.4 | 540 | | GET | page=5&page_size=25 | 249 | 0(0.00%) | 392 | 323 | 1367 | 370 | 0.4 | 530 | | GET | page=7&page_size=50 | 243 | 0(0.00%) | 392 | 319 | 997 | 360 | 0.4 | 530 | | GET | page=9&page_size=50 | 221 | 0(0.00%) | 391 | 330 | 658 | 360 | 0 | 520 | | GET | page=8&page_size=50 | 241 | 0(0.00%) | 388 | 317 | 1300 | 360 | 0.4 | 520 | | GET | page=5&page_size=50 | 204 | 0(0.00%) | 386 | 322 | 723 | 360 | 0.1 | 520 | | GET | page=2&page_size=25 | 259 | 0(0.00%) | 363 | 299 | 1124 | 340 | 0.4 | 520 | | GET | page=1&page_size=25 | 242 | 0(0.00%) | 373 | 296 | 1142 | 350 | 0.4 | 510 | | GET | page=4&page_size=25 | 238 | 0(0.00%) | 361 | 292 | 1171 | 340 | 0.3 | 500 | | GET | page=6&page_size=25 | 256 | 0(0.00%) | 365 | 312 | 687 | 350 | 0.3 | 480 | | GET | page=3&page_size=25 | 232 | 0(0.00%) | 356 | 302 | 752 | 340 | 0.9 | 480 | | GET | page=7&page_size=25 | 261 | 0(0.00%) | 362 | 310 | 590 | 350 | 0.1 | 470 | | GET | page=9&page_size=25 | 274 | 0(0.00%) | 304 | 254 | 797 | 280 | 0.3 | 430 | | GET | page=8&page_size=25 | 232 | 0(0.00%) | 300 | 253 | 644 | 280 | 0 | 420 | | GET | page=8&page_size=1 | 235 | 0(0.00%) | 228 | 191 | 1025 | 210 | 0 | 320 | | GET | page=3&page_size=1 | 255 | 0(0.00%) | 235 | 185 | 1106 | 220 | 0.2 | 310 | | GET | page=9&page_size=1 | 248 | 0(0.00%) | 234 | 184 | 1147 | 210 | 0.3 | 310 | | GET | page=5&page_size=1 | 247 | 0(0.00%) | 233 | 186 | 851 | 220 | 0.5 | 300 | | GET | page=7&page_size=1 | 231 | 0(0.00%) | 234 | 188 | 2104 | 220 | 0.3 | 290 | | GET | page=2&page_size=1 | 251 | 0(0.00%) | 226 | 187 | 972 | 220 | 0.4 | 290 | | GET | page=1&page_size=1 | 255 | 0(0.00%) | 223 | 186 | 461 | 210 | 0.3 | 290 | | GET | page=4&page_size=1 | 260 | 0(0.00%) | 222 | 188 | 400 | 210 | 0.3 | 290 | | GET | page=6&page_size=1 | 266 | 0(0.00%) | 221 | 187 | 373 | 220 | 0.1 | 280 |
|
...
This tests takes the course BerkeleyX/ColWri2.2x/1T2014 (~38,000 threads, ~40,000 comments) and then retrieves a single thread. This thread is then grouped by its comment_count. This test was run for 1 hour. As expected, because of pagination issues, threads with more comments are taking longer to return.
...