Repository navigation
Expand file tree
/
Copy pathtests.yaml
More file actions
765 lines (763 loc) · 25.4 KB
/
Copy pathtests.yaml
File metadata and controls
765 lines (763 loc) · 25.4 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
# Copyright 2026 The Agent Substrate Authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# The orchestrator deploys + tears down substrate + workloads around each
# test so runs don't pollute each other.
#
# Fields shared by every entry:
#
# type Required. Selects the runner: "locust" | "nighthawk-ingress".
# Each type's own fields are described in its section.
# targetCluster Required. Names a config file under
# /etc/orchestrator/target-clusters/ (e.g. `targetCluster:
# dev` -> /etc/orchestrator/target-clusters/dev.sh). The
# orchestrator copies that file into the cloned repo as
# .ate-dev-env.sh before the test runs and deletes it
# after, so different tests can target different clusters
# / projects from one CronJob run.
# sandboxClass Optional (default "gvisor"). Sandbox runtime for the
# benchmark WorkerPool: "gvisor" | "microvm".
# workerCount Optional (default 1). WorkerPool replica count for the
# deploy that runs before this test. The orchestrator
# redeploys workloads between tests, so each test can
# pick its own scale.
# actorMemory Optional (default "256Mi"). Memory limit on the
# benchmark ActorTemplates;
# raise it for suites that exercise glutton's WriteRAM path.
# actorCPU Optional (default "1"). CPU limit on the benchmark
# ActorTemplates. The scheduler books it against each
# worker's CPU capacity, so it caps the actors a worker
# hosts at once; lower it for dense suites.
# workerMemory Optional (default unset: the worker pods have no requests
# or limits). Memory request and limit on each WorkerPool
# pod. Use it to fix how many workers share a node: one per
# node needs more than half the node's allocatable memory.
# The limit, less what the worker keeps for itself, is
# the memory it reports as its actor capacity.
# workerWaitTimeout
# Optional (default 300, set by workloads/deploy.sh).
# Whole seconds to wait after the WorkerPool deploy for
# the worker pods to become ready. Raise it for
# large workerCount fleets whose image pulls or scheduling
# outrun the default.
# ateArgs Optional. List of strings appended verbatim to
# `ate-setup deploy ate-system` when this
# test's substrate is deployed (e.g. ["--atenet-dataplane=agentgateway"]).
# The deploy requires --credential-provider; set it here
# as one YAML string, '--credential-provider={"name":"k8s.io"}'
# (no shell quoting: the args are passed without a shell),
# or as ATE_CREDENTIAL_PROVIDER in .ate-dev-env.sh.
# isolate Optional. Isolate mode: installs substrate with the fake
# data plane (ate-setup deploy ate-system --fake-data-plane), which
# serves the WorkerPool without worker pods: fake-atelet
# takes atelet's place on every atelet node and answers each
# lifecycle call after a delay, and fake-workersync honors
# the pool with workerCount fake Workers on them, each
# reporting the capacity a real worker would.
# ate-api-server and Postgres can then be driven past what
# real nodes restore. The keys tune fake-atelet, through
# workloads/deploy.sh --fake-atelet-env: delay (default 0s),
# delayRun, delayRestore, delayCheckpoint,
# delayPromoteSnapshot, and delayTerminate, which
# override delay for one call, and mintActorCertificate
# (default true). A bare `isolate:` selects the mode with
# the defaults. Actors do not exist, so the entry's user
# class must send them no requests.
#
# ---- type: locust -----------------------------------------------------------
#
# Runs `runner.py -f <file> -t <duration> -u <users> --name <name>` in a
# Job, with --tag set to the build commit hash and --dest from the
# orchestrator's --dest flag. `file`, `duration` and `users` are required.
# Optional `flags` is a list of arbitrary strings appended to runner.py's
# argv; runner.py forwards them to locust verbatim.
#
# Optional `runnerCpu` (default "500m") and `runnerMemory` (default "512Mi")
# are the runner pod's requests. They are requests only, no limits, so the
# pod lands on a node with that much free and may burst beyond it. Raise
# them for boomer-driven classes at thousands of users: agentsession holds
# one goroutine per session and marshals up to 32Mi per in-flight ingest,
# so 1000+ sessions want on the order of 4 CPU and 8Gi on a node that has it.
#
# `file` is one path, or a comma-separated list of paths that locust reads as
# one test. Give the test first, then each load shape. A shape changes the
# load while the run continues; the shapes are in benchmarking/locust/shapes/.
#
# `--ladder users:duration[:trace_probability],...` with
# shapes/ladder_shape.py in `file` makes a sweep one run. The shape owns the
# load and the end of the run, thus locust ignores `-t` and `users` is the
# count before the first tick only. Keep `duration` at the total of the steps:
# the orchestrator sizes the Job timeout from it.
#
# A step of the ladder that gives a trace probability changes the sample rate
# when the step starts, and the new rate stays until another step changes it.
# `--trace-probability` in `flags` is the rate before the first such step.
# Thus a ladder with no probability in any step keeps the rate of the flag.
#
# A step can end in `@target_rps` (e.g. `200:3m@4000`), which caps the
# requests per second of the whole boomer worker from the step's start; 0
# removes the cap. Boomer users that pace themselves (resumepaused) honor it.
# Keep `users` at or above the rate times the p99 of one user iteration in
# seconds, or the step falls short of its rate. A resumepaused iteration is a
# resume plus a pause, so size it by the p99 of both together.
#
# An optional `controlPlane:` block shapes the control plane for a capacity
# run; pre_test applies it after the deploy. Its `apiServer:` section, which
# is required, sizes ate-api-server:
# cpu Required. CPU of each replica, set as both request and
# limit (e.g. 2 or "1500m").
# memory Optional. Memory request and limit of each replica.
# replicas Optional.
# poolMaxConns Optional. pgxpool max connections per replica, set
# through ateapi's env ConfigMap.
# An optional knob left unset keeps the installed value.
# Its optional `postgres:` section:
# stats Optional (default true). Load pg_stat_statements and add
# a postgres_exporter sidecar, scraped by monitoring.yaml.
# args Optional. Postgres settings as name=value (e.g.
# commit_delay=100), applied in the restart stats makes.
# The block leaves placement to the install. A capacity run sets
# --cordon-system-infra-components in ateArgs, which keeps ate-api-server and
# Postgres off the nodes that run the load, on the tainted pools that flag
# expects.
#
# ---- type: nighthawk-ingress --------------------------------------------------------
#
# Runs the Nighthawk adaptive router-capacity benchmark. `duration` is the
# job-wait budget only; the test knobs live in a `nighthawk-ingress:` block — see
# the "Configuration knobs" table in benchmarking/nighthawk-ingress/README.md.
# `workerCount` is required: the benchmark warms one actor per worker, so
# it is also the actor fleet size.
tests:
- name: glutton_baseline_1_user
type: locust
description: "Glutton baseline: 1 concurrent user for 1 minute"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--lifecycle-mode"
- "sync"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_baseline_1_user_pause
type: locust
description: "Glutton baseline (pause): 1 concurrent user for 1 minute"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--lifecycle-mode"
- "async"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
# Large-memory variants: same load shape as the 1-user baseline, but the
# GluttonUser fills the actor's RAM to --mem-target through the
# glutton WriteRAM API before the first suspend, so every suspend/resume
# cycle is measured against a realistically-sized actor. actorMemory =
# target + headroom (glutton itself and the guest). The golden snapshot
# and cold boot stay small: the working set exists from the first fill
# onward. --mem-churn re-randomizes 64Mi
# each cycle through an advancing window (WriteRAM rotate), so repeated
# suspends snapshot changing memory rather than a static set and the dirty
# window moves across the array. --mem-read walks the whole working set
# right after every resume, one byte per page, so under a demand-paged
# restore the GluttonReadRAM row is the cost of paging the previous
# snapshot's memory back in.
- name: glutton_mem_1gi_gvisor
type: locust
description: "Large-memory: 1Gi resident working set, gvisor"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
actorMemory: "1536Mi"
flags:
- "--lifecycle-mode"
- "sync"
- "--mem-target"
- "1Gi"
- "--mem-churn"
- "64Mi"
- "--mem-read"
- "all"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_mem_1gi_gvisor_pause
type: locust
description: "Large-memory (pause): 1Gi resident working set, gvisor"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
actorMemory: "1536Mi"
flags:
- "--lifecycle-mode"
- "async"
- "--mem-target"
- "1Gi"
- "--mem-churn"
- "64Mi"
- "--mem-read"
- "all"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_mem_1gi_microvm
type: locust
description: "Large-memory: 1Gi resident working set, microvm"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
sandboxClass: microvm
actorMemory: "1536Mi"
flags:
- "--mem-target"
- "1Gi"
- "--mem-churn"
- "64Mi"
- "--mem-read"
- "all"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_mem_2gi_gvisor
type: locust
description: "Large-memory: 2Gi resident working set, gvisor"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
actorMemory: "2560Mi"
flags:
- "--lifecycle-mode"
- "sync"
- "--mem-target"
- "2Gi"
- "--mem-churn"
- "64Mi"
- "--mem-read"
- "all"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_mem_2gi_gvisor_pause
type: locust
description: "Large-memory (pause): 2Gi resident working set, gvisor"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
actorMemory: "2560Mi"
flags:
- "--lifecycle-mode"
- "async"
- "--mem-target"
- "2Gi"
- "--mem-churn"
- "64Mi"
- "--mem-read"
- "all"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_mem_2gi_microvm
type: locust
description: "Large-memory: 2Gi resident working set, microvm"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 1
workerCount: 1
sandboxClass: microvm
actorMemory: "2560Mi"
flags:
- "--mem-target"
- "2Gi"
- "--mem-churn"
- "64Mi"
- "--mem-read"
- "all"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_baseline_5_users
type: locust
description: "Glutton baseline: 5 concurrent users for 1 minute"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 5
workerCount: 5
flags:
- "--lifecycle-mode"
- "sync"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_baseline_5_users_pause
type: locust
description: "Glutton baseline (pause): 5 concurrent users for 1 minute"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 5
workerCount: 5
flags:
- "--lifecycle-mode"
- "async"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_baseline_10_users
type: locust
description: "Glutton baseline: 10 concurrent users for 1 minute (doubles the number of workers/node)"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 10
workerCount: 10
flags:
- "--lifecycle-mode"
- "sync"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_baseline_10_users_pause
type: locust
description: "Glutton baseline (pause): 10 concurrent users for 1 minute"
targetCluster: dev
file: /app/tests/glutton.py
duration: 1m
users: 10
workerCount: 10
flags:
- "--lifecycle-mode"
- "async"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: glutton_oversubscribe_15_users
type: locust
description: "Glutton baseline: 15 concurrent users for 2 minutes (50% oversubscribed)"
targetCluster: dev
file: /app/tests/glutton.py
duration: 2m
users: 15
workerCount: 10
flags:
- "--min-wait-time"
- "0.5"
- "--max-wait-time"
- "1.0"
# The observability ladder of benchmarking/observability.md. Each entry
# runs on glutton, and thus on boomer: Python and gRPC hold each other at
# a high number of users, which would make the latency of the load
# generator part of the measurement.
- name: observability_s0_idle_floor
type: locust
description: "Observability S0: the telemetry of an idle control plane, 5 minutes"
targetCluster: dev
file: /app/tests/glutton.py,/app/shapes/ladder_shape.py
duration: 5m
users: 0
workerCount: 10
flags:
# The floor sends no request, thus locust writes no measurement row and
# the result of the run is in the telemetry counts only.
- "--allow-empty-stats"
- "--ladder"
- "0:5m"
- "--trace-probability"
- "0.1"
- name: observability_s1_user_sweep
type: locust
description: "Observability S1: user sweep 5 -> 10 -> 15, 3 minutes for each step"
targetCluster: dev
file: /app/tests/glutton.py,/app/shapes/ladder_shape.py
duration: 9m
users: 5
workerCount: 10
flags:
- "--ladder"
- "5:3m,10:3m,15:3m"
- "--trace-probability"
- "0.1"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: observability_s2_sample_rate_sweep
type: locust
description: "Observability S2: sample rate 0 -> 0.1 -> 1.0 at 10 users, 3 minutes for each step"
targetCluster: dev
file: /app/tests/glutton.py,/app/shapes/ladder_shape.py
duration: 9m
users: 10
workerCount: 10
flags:
- "--ladder"
- "10:3m:0,10:3m:0.1,10:3m:1.0"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: observability_s3_soak
type: locust
description: "Observability S3: 12 users for 10 minutes, 80% of the S1 maximum"
targetCluster: dev
file: /app/tests/glutton.py,/app/shapes/ladder_shape.py
duration: 10m
users: 12
workerCount: 10
flags:
- "--ladder"
- "12:10m"
- "--trace-probability"
- "0.1"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_data_baseline
type: locust
description: "DurDir data baseline: 1 concurrent user, 8 MiB file, explicit resume"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--lifecycle-mode"
- "sync"
- "--resume-mode"
- "explicit"
- "--durdir-file-size-bytes"
- "8388608"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_data_pause
type: locust
description: "DurDir data baseline (pause): 1 concurrent user, 8 MiB file, pause lifecycle"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--lifecycle-mode"
- "async"
- "--resume-mode"
- "explicit"
- "--durdir-file-size-bytes"
- "8388608"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_full_baseline
type: locust
description: "DurDir full baseline: 1 concurrent user, 8 MiB file, full memory restore comparison"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-full"
- "--lifecycle-mode"
- "sync"
- "--resume-mode"
- "explicit"
- "--durdir-file-size-bytes"
- "8388608"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_full_pause
type: locust
description: "DurDir full baseline (pause): 1 concurrent user, 8 MiB file, full memory restore comparison"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-full"
- "--lifecycle-mode"
- "async"
- "--resume-mode"
- "explicit"
- "--durdir-file-size-bytes"
- "8388608"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_implicit_resume
type: locust
description: "DurDir implicit resume: 1 concurrent user, 8 MiB file, traffic-triggered resume"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--resume-mode"
- "implicit"
- "--durdir-file-size-bytes"
- "8388608"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_size_5mb
type: locust
description: "DurDir file size sweep (5 MiB): 1 concurrent user, digest read mode"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--resume-mode"
- "explicit"
- "--durdir-read-mode"
- "digest"
- "--durdir-file-size-bytes"
- "5242880"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_size_10mb
type: locust
description: "DurDir file size sweep (10 MiB): 1 concurrent user, digest read mode"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--resume-mode"
- "explicit"
- "--durdir-read-mode"
- "digest"
- "--durdir-file-size-bytes"
- "10485760"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_size_64mb
type: locust
description: "DurDir file size sweep (64 MiB): 1 concurrent user, digest read mode"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--lifecycle-mode"
- "sync"
- "--resume-mode"
- "explicit"
- "--durdir-read-mode"
- "digest"
- "--durdir-file-size-bytes"
- "67108864"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: durdir_size_64mb_pause
type: locust
description: "DurDir file size sweep (64 MiB, pause): 1 concurrent user, digest read mode"
targetCluster: dev
file: /app/tests/durdir.py
duration: 1m
users: 1
workerCount: 1
flags:
- "--durdir-template"
- "glutton-durdir-data"
- "--lifecycle-mode"
- "async"
- "--resume-mode"
- "explicit"
- "--durdir-read-mode"
- "digest"
- "--durdir-file-size-bytes"
- "67108864"
- "--min-wait-time"
- "1.0"
- "--max-wait-time"
- "1.0"
- name: spawn_smoke_10_actors
type: locust
description: "Spawn benchmark smoke: 10 actors created with concurrency 2"
targetCluster: dev
file: /app/tests/spawn.py,/app/shapes/spawn_shape.py
duration: 5m
users: 1
workerCount: 10
flags:
- "--total-actors"
- "10"
- "--spawn-concurrency"
- "2"
- "--actor-deadline"
- "120"
- name: ingress_routercap_envoy_2cpu
description: "Router capacity: Nighthawk adaptive max-RPS search under a 25ms tail-latency SLO, Envoy at 2 CPUs"
type: nighthawk-ingress
targetCluster: dev
duration: 30m
workerCount: 50
nighthawk-ingress:
envoyCpu: 2
tailLatencySloMs: 25
- name: ingress_routercap_envoy_4cpu
description: "Router capacity: Nighthawk adaptive max-RPS search under a 25ms tail-latency SLO, Envoy at 4 CPUs"
type: nighthawk-ingress
targetCluster: dev
duration: 30m
workerCount: 50
nighthawk-ingress:
envoyCpu: 4
tailLatencySloMs: 25
- name: ingress_routercap_envoy_8cpu
description: "Router capacity: Nighthawk adaptive max-RPS search under a 25ms tail-latency SLO, Envoy at 8 CPUs"
type: nighthawk-ingress
targetCluster: dev
duration: 30m
workerCount: 50
nighthawk-ingress:
envoyCpu: 8
tailLatencySloMs: 25
- name: ingress_routercap_envoy_16cpu
description: "Router capacity: Nighthawk adaptive max-RPS search under a 25ms tail-latency SLO, Envoy at 16 CPUs"
type: nighthawk-ingress
targetCluster: dev
duration: 30m
workerCount: 50
nighthawk-ingress:
envoyCpu: 16
tailLatencySloMs: 25
# ate-api-server and Postgres cold resume ceiling on the fake data plane.
# Each user owns one actor: its first iteration creates the actor and
# makes it cold, in the step that adds the user, and every later one
# resumes it and pauses it again. fake-atelet answers each call, restore
# and checkpoint alike, after 150 ms on every atelet node, so a cycle
# spends 300 ms in the data plane, after minting the actor's certificate
# on each restore as ateom does. The 200 fake Workers hold 1000 actors
# each, and at 10m CPU an actor books next to nothing of a Worker's CPU,
# which a fake Worker reports as its node's allocatable, so placement
# never binds on any node size. resumepaused sends the actors no
# requests, which the fake data plane needs.
- name: ateapi_cold_resume_pause_fake_2cpu
type: locust
description: "ate-api-server cold resume on the fake data plane, ate-api-server at 2 CPUs: rate ladder 100-1500 cold resumes/s"
targetCluster: dev
file: /app/tests/resumepaused.py,/app/shapes/ladder_shape.py
duration: 23m
users: 1000
workerCount: 200
actorCPU: "10m"
runnerCpu: "8"
runnerMemory: "2Gi"
isolate:
delay: 150ms
ateArgs:
- "--cordon-system-infra-components"
controlPlane:
apiServer:
replicas: 1
cpu: 2
memory: 4Gi
poolMaxConns: 64
flags:
- "--ladder-spawn-rate"
- "1000"
- "--ladder"
- "1000:3m@100,1000:3m@200,1000:3m@400,1000:3m@600,1000:3m@800,1000:3m@1000,1000:3m@1500"
# Smoke for the cold resume setup above: the same fake data plane and
# ate-api-server shape at a load far below its ceiling, to check a change
# end to end in minutes. 50 users, each a cycle of about 300 ms, can make
# some 150 cold resumes/s, so both steps reach their rate.
- name: ateapi_cold_resume_pause_fake_smoke
type: locust
description: "ate-api-server cold resume smoke on the fake data plane, ate-api-server at 2 CPUs: 50 actors, 20 then 50 cold resumes/s"
targetCluster: dev
file: /app/tests/resumepaused.py,/app/shapes/ladder_shape.py
duration: 6m
users: 50
workerCount: 2
actorCPU: "10m"
isolate:
delay: 150ms
ateArgs:
- "--cordon-system-infra-components"
controlPlane:
apiServer:
replicas: 1
cpu: 2
memory: 4Gi
poolMaxConns: 64
flags:
- "--ladder-spawn-rate"
- "50"
- "--ladder"
- "50:2m@20,50:2m@50"