-
Notifications
You must be signed in to change notification settings - Fork 164
/
config-defaults.yaml
1201 lines (1001 loc) · 45 KB
/
config-defaults.yaml
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
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# Autoscaling settings
autoscaling_scale_down_enabled: "true"
autoscaling_buffer_cpu: "1m"
autoscaling_buffer_memory: "10Mi"
autoscaling_buffer_pods: "1"
cluster_autoscaler_cpu: "100m"
cluster_autoscaler_memory: "300Mi"
autoscaling_utilization_threshold: "1.0"
autoscaling_max_empty_bulk_delete: "10"
autoscaling_scale_down_unneeded_time: "2m"
autoscaling_unremovable_node_recheck_timeout: "5m"
# configure the log level for the autoscaler. Setting this to 4 gives enough
# verbosity to understand why a certain node pool is not picked for scheduling a
# pod.
autoscaling_autoscaler_log_level: "1"
# How long to wait for pod eviction when scaling down.
{{if eq .Cluster.Environment "production"}}
cluster_autoscaler_max_pod_eviction_time: "1h"
{{else}}
cluster_autoscaler_max_pod_eviction_time: "3h"
{{end}}
# Override terminationGracePeriodSeconds when evicting pods for scale down, if the pods' value is higher than this one
cluster_autoscaler_max_graceful_termination_sec: "1209600" # 2 weeks
# Prevent CA lock-ups caused by large amounts of pending pods. Can be disabled completely by setting to 0.
cluster_autoscaler_max_usnchedulable_pods_considered: "1000"
# karpenter settings
# DO NOT SET TO FALSE IF THE CLUSTER HAS KARPENTER POOLS OR NODES. REFER TO TEAPOT DOCS FOR HOW TO ROLLBACK KARPENTER
# https://teapot.docs.zalando.net/howtos/karpenter-operations/
karpenter_pools_enabled: "false"
karpenter_controller_cpu: "25m"
karpenter_controller_memory: "256Mi"
# set log level of karpenter: error|debug
karpenter_log_level: "error"
# restrict the maximum number of pods for karpenter nodes
karpenter_max_pods_per_node: "32"
#
# Karpenter version for controlling roll-out, can be "current" or "legacy"
# current => 0.37.0-main-26.patched
# legacy => 0.36.2-main-25.patched
karpenter_version: "current"
# ALB config created by kube-aws-ingress-controller
kube_aws_ingress_controller_ssl_policy: "ELBSecurityPolicy-TLS-1-2-2017-01"
kube_aws_ingress_controller_idle_timeout: "1m"
kube_aws_ingress_controller_deregistration_delay_timeout: "10s"
# allow using NLBs for ingress
# This opens skipper-ingress ports 9998 and 9999 on all worker nodes
kube_aws_ingress_controller_nlb_enabled: "true"
kube_aws_ingress_controller_nlb_cross_zone: "true"
kube_aws_ingress_controller_nlb_zone_affinity: "any_availability_zone"
kube_aws_ingress_controller_cert_polling_interval: "2m"
# sets the default LB type: "network" or "application" are valid choices (overwritten by nlb_switch)
kube_aws_ingress_default_lb_type: "application"
# cert filter
kube_aws_ingress_controller_cert_filter_tag: "kubernetes=enabled"
# ALB to NLB switch
# "pre":
# - kube-ingress-aws-controller will configure NLB to forward HTTPS requests
# to skipper-ingress on port 9999 and HTTP requests to port 9998
# - skipper-ingress on port 9999 will add X-Forwarded-For=<client IP> and X-Forwarded-Proto=https headers to requests from NLB
# - skipper-ingress on port 9998 will reply with 308 Permanent Redirect to https
#
# "exec": same as "pre" and kube-ingress-aws-controller will default to NLB
# after removing it you need to set kube_aws_ingress_default_lb_type: "network" above and cleanup skipper and ingress-ctl deployment.yaml
nlb_switch: "exec"
# skipper ingress settings
skipper_ingress_target_average_utilization_cpu: "60"
skipper_ingress_target_average_utilization_memory: "80"
skipper_ingress_hpa_scale_down_wait: "600"
skipper_ingress_hpa_scale_up_max_perc: "100"
{{if eq .Cluster.Environment "production"}}
skipper_ingress_min_replicas: "3"
skipper_ingress_max_replicas: "300"
{{else}}
skipper_ingress_min_replicas: "2"
skipper_ingress_max_replicas: "50"
{{end}}
skipper_ingress_cpu: "1000m"
skipper_ingress_memory: "1500Mi"
skipper_topology_spread_enabled: "true"
skipper_topology_spread_timeout: "7m"
# PHC
skipper_ingress_health_check_options: "period=10s,min-requests=10,min-drop-probability=0.05,max-drop-probability=0.9,max-unhealthy-endpoints-ratio=0.9"
# Enables deployment of canary version
skipper_ingress_canary_enabled: "true"
skipper_ingress_test_single_pod: "false"
skipper_canary_controller_enabled: "false"
# When set to true (and dedicated node pool for skipper is also true) the
# daemonset overhead will be subtracted from the cpu settings such
# that skipper will perfectly fit on the node.
{{if eq .Cluster.Environment "e2e"}}
skipper_ingress_binpack: "true"
{{else}}
skipper_ingress_binpack: "false"
{{end}}
# skipper node-pool
enable_dedicate_nodepool_skipper: "true"
{{if eq .Cluster.Environment "e2e"}}
skipper_attach_only_to_skipper_node_pool: "false"
{{else}}
skipper_attach_only_to_skipper_node_pool: "true"
{{end}}
skipper_suppress_route_update_logs: "true"
skipper_validate_query: "true"
skipper_validate_query_log: "false"
skipper_default_filters: 'disableAccessLog(2,3,404,429) -> fifo(2000,20,"1s")'
# skipper_default_filters_authentication defines filters that implement default request authentication
skipper_default_filters_authentication: ''
skipper_default_filters_append: 'stateBagToTag("auth-user", "client.uid")'
skipper_disabled_filters: "static,bearerinjector,setRequestHeaderFromSecret,basicAuth"
skipper_lua_sources: "file"
skipper_edit_route_placeholders: ""
skipper_ingress_inline_routes: ""
skipper_ingress_refuse_payload: ""
skipper_endpointslices_enabled: "true"
skipper_kubernetes_annotation_predicates: ''
skipper_kubernetes_east_west_range_annotation_predicates: ''
skipper_compress_encodings: "gzip,deflate,br"
# Adds "start" label to each prometheus counter with the value of counter creation timestamp as unix nanoseconds
skipper_prometheus_start_label_enabled: "true"
# skipper profiling settings, 0 keeps default, <0 disable, >0 enable with value
# https://pkg.go.dev/runtime@master#SetBlockProfileRate
# https://pkg.go.dev/runtime@master#SetMutexProfileFraction
skipper_block_profile_rate: 0
skipper_mutex_profile_fraction: 0
skipper_memory_profile_rate: 0
# skipper backend timeout defaults
skipper_expect_continue_timeout_backend: "30s"
skipper_keepalive_backend: "30s"
skipper_max_idle_connection_backend: "0"
skipper_response_header_timeout_backend: "1m"
skipper_timeout_backend: "1s"
skipper_tls_timeout_backend: "3s"
skipper_close_idle_conns_period: "20s"
# skipper server timeout defaults
skipper_read_timeout_server: "5m"
skipper_write_timeout_server: "0"
# skipper server keepalive defaults
skipper_keepalive_server: "0"
skipper_keepalive_requests_server: "0"
# skipper startup settings
{{if eq .Cluster.Environment "production"}}
skipper_readiness_init_delay_seconds: 60
skipper_liveness_init_delay_seconds: 30
{{else}}
skipper_readiness_init_delay_seconds: 1
skipper_liveness_init_delay_seconds: 30
{{end}}
# skipper termination settings
# (10s LB healthcheck interval) * (3 unhealthy threshold + margin of 1 interval)
skipper_wait_for_healthcheck_interval: "40s"
# (350s of fixed NLB connection idle timeout) + (margin of 2s)
skipper_idle_timeout_server: "352s"
# wait long enough for LB to detect unhealthy node and all connections become idle,
# i.e. skipper_wait_for_healthcheck_interval + skipper_idle_timeout_server
skipper_termination_grace_period: "392"
# skipper redis settings
enable_dedicate_nodepool_skipper_redis: "false"
skipper_redis_cpu: "100m"
skipper_redis_memory: "512Mi"
skipper_redis_dial_timeout: "25ms"
skipper_redis_pool_timeout: "250ms"
skipper_redis_read_timeout: "25ms"
skipper_redis_write_timeout: "25ms"
skipper_redis_min_conns: 25
skipper_redis_max_conns: 100
skipper_ingress_redis_swarm_enabled: "true"
skipper_ingress_redis_swim_enabled: "false"
skipper_ingress_redis_target_average_utilization_cpu: "30"
skipper_ingress_redis_target_average_utilization_memory: "60"
skipper_ingress_redis_min_replicas: "1"
skipper_ingress_redis_max_replicas: "100"
skipper_ingress_redis_cluster_scaling_schedules: ""
skipper_ingress_redis_hpa_scale_down_wait: "600"
skipper_cluster_ratelimit_max_group_shards: 1
# webhook protection poc
skipper_lua_scripts_enabled: ""
## datadome poc
datadome_api_key: ""
## kasada poc
kasada_api_key: ""
#
# skipper routesrv settings
#
# skipper_routesrv_enabled is a three state switch:
# - "false" - routesrv deployment is removed, skipper uses own k8s dataclient
# - "pre" - routesrv is deployed, skipper uses own k8s dataclient
# - "exec" - routesrv is deployed, skipper uses routesrv
skipper_routesrv_enabled: "exec"
skipper_routesrv_memory: "1Gi"
{{if eq .Cluster.Environment "production"}}
skipper_routesrv_cpu: "1000m"
skipper_routesrv_min_replicas: 2
{{else}}
skipper_routesrv_cpu: "100m"
skipper_routesrv_min_replicas: 1
{{end}}
skipper_routesrv_node_affinity_enabled: "false"
skipper_routesrv_max_replicas: 10
skipper_routesrv_target_average_utilization_cpu: "80"
skipper_routesrv_target_average_utilization_memory: "80"
skipper_ingress_routesrv_scaling_schedules: ""
skipper_routesrv_log_level: "INFO"
# skipper-ingress component pod-deletion-cost-controller
skipper_pod_deletion_cost_controller_memory: 200Mi
skipper_pod_deletion_cost_controller_cpu: 50m
# vpa min values
skipper_pod_deletion_cost_controller_memory_min: 128Mi
skipper_pod_deletion_cost_controller_cpu_min: 25m
# klog style -v=0
skipper_pod_deletion_cost_controller_log_v: "0"
# pod informer sync config
skipper_pod_deletion_cost_controller_poll_interval: "10s"
skipper_pod_deletion_cost_controller_poll_timeout: "60s"
# enable re-sync
skipper_pod_deletion_cost_controller_resync_enable: "true"
skipper_pod_deletion_cost_controller_resync_interval: "1h"
# polarsignals - only enabled for testing teapot
polarsignals_enabled: "false"
# Kube-Metrics-Adapter
## Scheduled scaling metrics: ramp up/down over this period of time
kube_metrics_adapter_default_scaling_window: "10m"
## Scheduled scaling metrics: number of steps to scale. 5 allows at
## least 20% of change between each schedule and enough change to
## trigger the HPA.
kube_metrics_adapter_scaling_schedule_ramp_steps: "5"
## ZMON KairosDB URL
zmon_kairosdb_url: "https://data-service.zmon.zalan.do/kairosdb-proxy"
## Nakadi URL (for the stats API)
nakadi_url: ""
# enable temporary logging of ingress.cluster.local names
# used to find services for which it's being used.
skipper_eastwest_dns_log_enabled: "false"
# if enabled adds port 8080 as svc port to eastwest svc
skipper_ingress_eastwest_additional_port: "false"
# skipper tcp lifo
# See: https://opensource.zalando.com/skipper/operation/operation/#tcp-lifo
skipper_enable_tcp_queue: "true" # TODO(sszuecs): cleanup candidate to reduce amount of branches in deployment
skipper_expected_bytes_per_request: "51200"
skipper_max_tcp_listener_concurrency: "-1"
skipper_max_tcp_listener_queue: "-1"
# opentracing
skipper_ingress_opentracing_excluded_proxy_tags: ""
skipper_ingress_opentracing_backend_name_tag: "true"
skipper_opentracing_disable_filter_spans: "true"
# lightstep
skipper_ingress_tracing_buffer: "32768"
skipper_ingress_lightstep_grpc_max_msg_size: 16384000
skipper_ingress_lightstep_min_period: "500ms"
skipper_ingress_lightstep_max_period: "2500ms"
skipper_ingress_lightstep_max_log_key_len: 20
skipper_ingress_lightstep_max_log_value_len: 128
skipper_ingress_lightstep_max_logs_per_span: 20
skipper_ingress_lightstep_propagators: "lightstep"
# set to "log-events" to enable
skipper_ingress_lightstep_log_events: ""
lightstep_token: ""
tracing_collector_host: "tracing.platform-infrastructure.zalan.do"
# skipper_serve_method_metric sets the flag -serve-method-metric. It
# defines if the http method is included in the dimension
# of the skipper_serve_host_duration_seconds_bucket metric.
skipper_serve_method_metric: "false"
# skipper_serve_status_code_metric sets the flag -serve-status-code-metric. It
# defines if the http response status code is included in the dimension
# of the skipper_serve_host_duration_seconds_bucket metric.
skipper_serve_status_code_metric: "false"
# disabled|provisioned|enabled routegroup validation via skipper webhook
# can be one of disabled|provisioned|enabled
routegroups_validation: "enabled"
# disabled|enabled ingress validation via skipper webhook
ingresses_validation: "enabled"
# tokeninfo
{{if eq .Cluster.Environment "production"}}
# production|bridge|disabled
skipper_local_tokeninfo: "production"
{{else}}
# production|bridge|disabled
skipper_local_tokeninfo: "bridge"
{{end}}
# tokeninfo cache
# integer, non-zero value enables tokeninfo cache and sets the maximum number of cached tokens
skipper_tokeninfo_cache_size: 1000
# duration, non-zero value limits the lifetime of a cached tokeninfo
skipper_tokeninfo_cache_ttl: "30s"
# oauth2 UI login - grant flow
{{if eq .Cluster.Environment "e2e"}}
skipper_oauth2_ui_login: "false"
skipper_ingress_encryption_key: ""
{{else}}
skipper_oauth2_ui_login: "true"
{{end}}
# Comma-separated list of tokeninfo keys to retain
skipper_oauth2_ui_login_tokeninfo_keys: ""
# ClusterScalingSchedules
# One or multiple cluster scaling schedules can be configured as a
# comma-separated list of <cluster schedule name>=<target value> pairs.
# E.g. to configure "schedule1" cluster schedule with a target value of 3 and "schedule2" with a target value of 5 set
# skipper_cluster_scaling_schedules: "schedule1=3,schedule2=5"
skipper_cluster_scaling_schedules: ""
# Skipper Ingress/RouteGroup backend traffic split algorithm: traffic-predicate or traffic-segment-predicate
skipper_ingress_backend_traffic_algorithm: "traffic-segment-predicate"
# TODO: after a while we can remove this and hardcode (2023-06-30)
skipper_ingress_default_lb_algorithm: "powerOfRandomNChoices"
skipper_ingress_disable_catchall_routes: "true"
# Original RouteGroup CRD does not require hosts but we explicitly want it specified in our infrastructure
skipper_ingress_routegroup_crd_require_hosts: "true"
# Set defaults values that would enable Open Policy Agent in a skipper filter
skipper_open_policy_agent_enabled: "false"
skipper_open_policy_agent_styra_token: ""
#
# FabricGateway controller config
#
# fabric_gateway_controller_enabled:
# - false: scales controller to zero replicas
# - true: runs the controller
#
fabric_gateway_controller_enabled: "true"
fabric_gateway_controller_cpu: "50m"
fabric_gateway_controller_memory: "150Mi"
fabric_gateway_controller_allow_all_filters: "false"
fabric_gateway_controller_snapshots_history_limit: "3"
fabric_gateway_controller_enable_versioning: "true"
fabric_gateway_controller_ssl_policy: ""
fabric_gateway_controller_log_level: "INFO"
# kube-api-server settings
# EventRateLimit admission plugin configuration
# Reference: https://kubernetes.io/docs/reference/access-authn-authz/admission-controllers/#eventratelimit
event_rate_limit_enable: "true"
event_rate_limit_config_qps: "500"
event_rate_limit_config_burst: "1000"
# cadvisor settings
cadvisor_cpu: "150m"
cadvisor_memory: "150Mi"
cadvisor_profiling_enabled: "false"
cadvisor_enabled: "false"
# settings for enabling the kubelet-summary-metrics proxy and prometheus metric
# collection.
kubelet_summary_metrics_enabled: "true"
kubelet_summary_metrics_cpu: "100m"
kubelet_summary_metrics_memory: "512Mi"
kubelet_summary_metrics_hpa_max_replicas: "10"
kubelet_summary_metrics_hpa_cpu_target: "80"
kubelet_summary_metrics_hpa_memory_target: "80"
# enable collection of cadvisor metrics from the kubelet cadvisor endpoint.
kubelet_cadvisor_enabled: "true"
# node exporter settings
node_exporter_cpu: "20m"
node_exporter_memory: "75Mi"
node_exporter_experimental_metrics: "false"
# kube-proxy settings
kube_proxy_cpu: "50m"
kube_proxy_memory_request: "200Mi"
kube_proxy_memory_limit: "200Mi"
kube_proxy_sync_period: "15m0s"
kube_proxy_verbose_level: "2"
# flannel settings
flannel_cpu: "25m"
flannel_memory_request: "100Mi"
flannel_memory_limit: "100Mi"
flannel_awaiter_cpu: "25m"
flannel_awaiter_memory_request: "50Mi"
flannel_awaiter_memory_limit: "50Mi"
# nvidia device plugin
nvidia_device_plugin_cpu: "10m"
nvidia_device_plugin_memory: "50Mi"
# nvidia dcgm exporter
nvidia_dcgm_exporter_enabled: "false"
nvidia_dcgm_exporter_cpu: "10m"
nvidia_dcgm_exporter_memory: "200Mi"
# AWS EFA device plugin
aws_efa_device_plugin_enabled: "false"
aws_efa_device_plugin_cpu: "10m"
aws_efa_device_plugin_memory: "20Mi"
# static egress controller settings
static_egress_controller_enabled: "true"
# journald reader settings
journald_reader_enabled: "true"
journald_reader_cpu: "1m"
journald_reader_memory: "30Mi"
# Logging settings
logging_s3_bucket: "zalando-logging-{{ .Cluster.InfrastructureAccount | getAWSAccountID}}-{{ .Cluster.Region }}"
scalyr_team_token: ""
log_destination_infra: "scalyr/stups"
log_destination_both: "scalyr/main+stups"
log_destination_local: "scalyr/main"
# Central bucket to keep logging infrastructure logs
logging_infrastructure_s3_bucket: ""
vpa_cpu: "200m"
vpa_mem: "500Mi"
prometheus_cpu: "1000m"
prometheus_mem: "4Gi"
prometheus_mem_min: "2Gi"
prometheus_cpu_min: "0"
prometheus_retention_size: "49GB" # one GB less than prometheus' PVC
prometheus_retention_time: "1d"
prometheus_storage_size: "50Gi"
# Upstream defaults are too aggressive:
# https://prometheus.io/docs/prometheus/latest/configuration/configuration/#remote_write
prometheus_remote_write: "disabled"
# Maximum time a sample will wait in buffer.
prometheus_remote_batch_send_deadline: "30s"
# Initial retry delay. Gets doubled for every retry.
prometheus_remote_min_backoff: "3s"
# Maximum retry delay.
prometheus_remote_max_backoff: "10s"
# Comma-separated list of user ids allowed to access Prometheus UI
prometheus_ui_users: ""
metrics_service_cpu: "100m"
metrics_service_mem_max: "4Gi"
metrics_server_metric_resolution: "15s"
kube_aws_iam_controller_cpu: "5m"
kube_aws_iam_controller_mem: "50Mi"
kube_state_metrics_cpu: "100m"
kube_state_metrics_mem_max: "4Gi"
kube_state_metrics_mem_min: "120Mi"
kubernetes_lifecycle_metrics_mem_max: "4Gi"
kubernetes_lifecycle_metrics_mem_min: "120Mi"
kube_node_ready_controller_cpu: "50m"
kube_node_ready_controller_memory: "200Mi"
# Enable deployment of aws-cloud-controller-manager
aws_cloud_controller_manager_enabled: "true"
aws_cloud_controller_manager_cpu: "125m"
aws_cloud_controller_manager_memory: "512Mi"
# Kubernetes Downscaler (for non-production clusters)
{{if eq .Cluster.Environment "test"}}
downscaler_default_uptime: "Mon-Fri 07:30-20:30 Europe/Berlin"
downscaler_default_downtime: "never"
downscaler_enabled: "true"
{{else if eq .Cluster.Environment "e2e"}}
downscaler_default_uptime: "always"
downscaler_default_downtime: "never"
downscaler_enabled: "true"
{{else}}
downscaler_default_uptime: "always"
downscaler_default_downtime: "never"
downscaler_enabled: "false"
{{end}}
# HPA settings (defaults from https://kubernetes.io/docs/reference/command-line-tools-reference/kube-controller-manager/)
horizontal_pod_autoscaler_sync_period: "30s"
horizontal_pod_autoscaler_tolerance: "0.1"
horizontal_pod_downscale_stabilization: "5m0s"
# Vertical pod autoscaler version for controlling roll-out, can be "current" or "legacy"
# current => v1.0.0-internal.20
# legacy => v0.12.0-internal.19
vertical_pod_autoscaler_version: "current"
# Cluster update settings
{{if eq .Cluster.Environment "production"}}
drain_grace_period: "6h"
drain_min_pod_lifetime: "72h"
drain_min_healthy_sibling_lifetime: "1h"
drain_min_unhealthy_sibling_lifetime: "6h"
drain_force_evict_interval: "5m"
node_update_prepare_replacement_node: "true"
{{else}}
drain_grace_period: "2h"
drain_min_pod_lifetime: "8h"
drain_min_healthy_sibling_lifetime: "1h"
drain_min_unhealthy_sibling_lifetime: "1h"
drain_force_evict_interval: "5m"
node_update_prepare_replacement_node: "false" # don't wait for a replacement instance for on-demand pools in test clusters
{{end}}
# add NoSchedule taints to nodes being replaced
decommission_node_no_schedule_taint: "true"
# Teapot admission controller
teapot_admission_controller_default_cpu_request: "25m"
teapot_admission_controller_default_memory_request: "100Mi"
teapot_admission_controller_process_resources: "true"
teapot_admission_controller_process_resource_ephemeral_storage: "false"
teapot_admission_controller_default_ephemeral_storage_request: "25Mi"
teapot_admission_controller_max_ephemeral_storage_request: "5Gi"
teapot_admission_controller_application_min_creation_time: "2019-06-03T12:00:00Z"
teapot_admission_controller_ndots: "2"
teapot_admission_controller_inject_environment_variables: "true"
{{ if eq .Cluster.Environment "test" }}
teapot_admission_controller_inject_node_options_environment_variable: "true"
{{ else }}
teapot_admission_controller_inject_node_options_environment_variable: "false"
{{ end }}
teapot_admission_controller_deployment_default_max_surge: "5%"
teapot_admission_controller_deployment_default_max_unavailable: "1"
teapot_admission_controller_inject_aws_waiter: "true"
teapot_admission_controller_parent_resource_hash: "true"
## Defaults are set per-cluster
teapot_admission_controller_check_daemonset_resources: "true"
teapot_admission_controller_daemonset_reserved_cpu: "8"
teapot_admission_controller_daemonset_reserved_memory: "64Gi"
kubelet_system_reserved_cpu: "100m"
kubelet_system_reserved_memory: "164Mi"
kubelet_kube_reserved_cpu: "100m"
kubelet_kube_reserved_memory: "282Mi"
kubelet_image_gc_high_threshold: 50
kubelet_image_gc_low_threshold: 40
{{if eq .Cluster.Environment "production"}}
teapot_admission_controller_validate_application_label: "true"
teapot_admission_controller_validate_base_images: "true"
# Check container image compliance in production clusters. Be careful when thinking about changing this: Setting it to
# false will allow any container image to run in production clusters.
#
# If you are seeing issues with "docker-meta" check the next config field to allow-list certain namespaces.
teapot_admission_controller_validate_pod_images: "true"
# If you are seeing issues with the container image compliance checker dependency "docker-meta" you can designate
# a subset of namespaces to be allowed regardless with a regular expression on the namespace, e.g.:
#
# if docker-meta is down, do not reject container images running in `kube-system`
# teapot_admission_controller_validate_pod_images_soft_fail_namespaces: "^kube-system$"
teapot_admission_controller_validate_pod_template_resources: "true"
teapot_admission_controller_preemption_enabled: "true"
teapot_admission_controller_postgresql_delete_protection_enabled: "true"
teapot_admission_controller_namespace_delete_protection_enabled: "true"
{{else if eq .Cluster.Environment "e2e"}}
teapot_admission_controller_validate_application_label: "false"
teapot_admission_controller_validate_base_images: "false"
# Check container image compliance in e2e clusters. There are some exceptions to allow the e2e test suite to run.
teapot_admission_controller_validate_pod_images: "true"
teapot_admission_controller_validate_pod_template_resources: "false"
teapot_admission_controller_preemption_enabled: "true"
teapot_admission_controller_postgresql_delete_protection_enabled: "false"
teapot_admission_controller_namespace_delete_protection_enabled: "false"
{{else}}
teapot_admission_controller_validate_application_label: "false"
teapot_admission_controller_validate_base_images: "false"
# Do not check container image compliance in test clusters.
teapot_admission_controller_validate_pod_images: "false"
teapot_admission_controller_validate_pod_template_resources: "true"
teapot_admission_controller_preemption_enabled: "false"
teapot_admission_controller_postgresql_delete_protection_enabled: "false"
teapot_admission_controller_namespace_delete_protection_enabled: "false"
{{end}}
# Enable automatic replacement of vanity images with ECR images
teapot_admission_controller_resolve_vanity_images: "true"
{{if eq .Cluster.Environment "e2e"}}
teapot_admission_controller_ignore_namespaces: "^kube-system|((downward-api|kubectl|projected|statefulset|pod-network|scope-selectors|resourcequota|limitrange|sysctl|node-tests|e2e-kubelet-etc-hosts|csiinlinevolumes|job|dns)-.*)$"
teapot_admission_controller_crd_ensure_no_resources_on_delete: "false"
{{else}}
teapot_admission_controller_ignore_namespaces: "^kube-system$"
teapot_admission_controller_crd_ensure_no_resources_on_delete: "true"
{{end}}
# Enable kube-node-ready-controller and node-not-ready taint
teapot_admission_controller_node_not_ready_taint: "true"
# Some third-party controllers use API groups that look like they belong to Kubernetes resources. Explicitly allow them anyway.
teapot_admission_controller_crd_role_provisioning_allowed_api_groups: "flink.k8s.io"
teapot_admission_controller_topology_spread: optin
teapot_admission_controller_topology_spread_timeout: 7m
# Enable and configure runtime-policy annotation
{{if eq .Cluster.Environment "production"}}
teapot_admission_controller_runtime_policy_default: "require-on-demand"
{{else}}
teapot_admission_controller_runtime_policy_default: "allow-spot"
{{end}}
# Enforce a certain policy (<empty>|allow-spot|require-on-demand) for a cluster,
# leave empty for falling back to the default.
teapot_admission_controller_runtime_policy_enforced: ""
# Enable hard spot assignment. Only relevant when node_lifecycle_provider=zalando
teapot_admission_controller_runtime_policy_spot_hard_assignment: "false"
# Enable experimental rescheduling of spot nodes after spot decommission
spot_node_rescheduler: "false"
spot_node_rescheduler_memory: "348Mi"
spot_node_rescheduler_cpu: "50m"
# Enable and configure prevent scale down annotation
teapot_admission_controller_prevent_scale_down_enabled: "true"
{{if eq .Cluster.Environment "production"}}
teapot_admission_controller_prevent_scale_down_allowed: "true"
{{else}}
teapot_admission_controller_prevent_scale_down_allowed: "false"
{{end}}
teapot_admission_controller_log4j_format_msg_no_lookups: "true"
teapot_admission_controller_graceful_termination: "true"
# toggle that prevents active configmaps from being deleted
{{if eq .Cluster.Environment "e2e"}}
# allow deletion of active configmaps in e2e clusters because some tests rely on it
teapot_admission_controller_configmap_deletion_protection_enabled: "false"
teapot_admission_controller_configmap_deletion_protection_factories_enabled: "false"
{{else}}
# prevent deletion of active configmaps in all test and production clusters
teapot_admission_controller_configmap_deletion_protection_enabled: "true"
teapot_admission_controller_configmap_deletion_protection_factories_enabled: "true"
{{end}}
# enable the rolebinding admission-controller webhook which validates rolebindings and clusterrolebindings
teapot_admission_controller_enable_rolebinding_webhook: "true"
# enable the generic deny-all admission webhook which rejects all requests it receives
teapot_admission_controller_enable_write_protection_webhook: "false"
# configure the behaviour of the deny-all admission webhook, `true` blocks everything, `false` allows everything
teapot_admission_controller_prevent_write_operations: "false"
# Enable and configure Pod Security Policy rules implemented in admission-controller.
teapot_admission_controller_pod_security_policy_enabled: "true"
# comma separated list of service accounts that are allowed to use privileged
# pod security policy rules. Format: `<namespace>_<service-account-name>`
{{ if eq .Cluster.Environment "e2e" }}
teapot_admission_controller_pod_security_policy_privileged_service_accounts: "psp-privileged-zalando_privileged-sa,psp-privileged-deployment-zalando_privileged-sa"
{{ else }}
teapot_admission_controller_pod_security_policy_privileged_service_accounts: ""
{{ end }}
teapot_admission_controller_pod_security_policy_privileged_allow_privilege_escalation: "false"
# comma separated list of restricted or privileged capabilities that are allowed
# to be used by pods.
teapot_admission_controller_pod_security_policy_additional_restricted_capabilities: ""
teapot_admission_controller_pod_security_policy_privileged_capabilities: ""
# Prevent the use of a particular AZ as much as possible
blocked_availability_zone: ""
# etcd cluster
etcd_stack_name: "etcd-cluster-etcd"
# comma separated list of DNS record prefixes which will be prefixed to the
# hosted zone of the account. We allow multiple prefixes for the purpose of
# migration.
etcd_dns_record_prefixes: "etcd-server.{{.Cluster.Region}}"
{{if eq .Cluster.Environment "production"}}
etcd_instance_count: "5"
etcd_instance_type: "m5.large"
{{else}}
etcd_instance_count: "3"
etcd_instance_type: "t3.medium"
{{end}}
etcd_scalyr_key: ""
etcd_ami: {{ amiID "zalando-ubuntu-etcd-production-v3.5.13-amd64-main-34" "861068367966"}}
cluster_dns: "coredns"
coredns_log_svc_names: "true"
coredns_log_forward: "false"
# max concurrency for upstream (AWS VPC) DNS server
#
# AWS VPC DNS server has a limit of 1024 qps before packets are dropped.
# This setting is tuned to allow a buffer compared to the normal DNS QPS in our
# clusters and prevent CoreDNS from running out of memory in case of spikes.
coredns_max_upstream_concurrency: 2000 # 0 means there is no concurrency limits
tracing_coredns_route_traces_to_local_zone: "false"
tracing_coredns_global_traces_endpoint: ""
tracing_coredns_local_zone_traces_endpoint: ""
# Kubernetes on Ubuntu AMI to use
# note this configuration uses the [amiID][0] function. It returns the
# AMI id given the image name and the Image AWS account owner.
#
# [0]: https://github.com/zalando-incubator/cluster-lifecycle-manager/blob/8a9bd1cb2d094038a9e23e646421f8146b48886a/provisioner/template.go#L116
kuberuntu_image_v1_31_jammy_amd64: {{ amiID "zalando-ubuntu-jammy-22.04-kubernetes-production-v1.31.4-amd64-master-359" "861068367966" }}
kuberuntu_image_v1_31_jammy_arm64: {{ amiID "zalando-ubuntu-jammy-22.04-kubernetes-production-v1.31.4-arm64-master-359" "861068367966" }}
# Which distro from the previous config items should be used. Valid options are only `jammy` for now. Can be set for each node pool.
kuberuntu_distro_master: "jammy"
kuberuntu_distro_worker: "jammy"
# Feature toggle for auditing events
audit_pod_events: "true"
# Whether to log audit event for node changes by kube-controller-manager. Useful
# to see if kube-controller-manager is setting overlapping podCIDRs.
audit_kube_controller_manager_node_changes: "false"
{{if eq .Cluster.Environment "production"}}
audittrail_url: "https://audittrail.cloud.zalando.com"
{{else}}
audittrail_url: ""
{{end}}
audittrail_nakadi_url: ""
audittrail_root_account_role: ""
audittrail_adapter_cpu: "50m"
audittrail_adapter_memory: "200Mi"
audittrail_adapter_timeout: "2s"
# When enabled, any read-only events are added to the metrics, but are dropped
# before being sent to audittrail-api.
# When this is set to true, `auditlog_read_access` can safely be set true
# without overloading audittrail-api.
audittrail_adapter_drop_audittrail_api_read_only: "true"
audit_webhook_batch_max_size: "250"
kube2iam_cpu: "25m"
kube2iam_memory: "100Mi"
# configure whether kube2iam should only run on worker nodes.
kube2iam_worker_only: "true"
# CIDR configuration for nodes and pods
# Changing this will change the number of nodes and pods we can schedule in the
# cluster: https://cloud.google.com/kubernetes-engine/docs/how-to/flexible-pod-cidr
{{if eq .Cluster.Environment "production"}}
node_cidr_mask_size: "25"
{{else}}
node_cidr_mask_size: "24"
{{end}}
# How many nodes to keep reserved (e.g. to allow for increasing the node_cidr_mask_size).
# Note that this only affects CA settings, someone can still scale up the ASGs manually.
reserved_nodes: "5"
# How much extra capacity to add when calculating the maximum number of pods per node. This can be increased if some
# pods don't consume the IP space on the node, but it's fairly dangerous since it has to be absolutely correct. Use
# in emergencies only, and pay extra attention when adding, removing or updating daemonsets.
node_max_pods_extra_capacity: "0"
# maximum number of PIDs allowed to be allocated per pod
pod_max_pids: "4096"
# the cpu management policy which should be used by the kubelet
cpu_manager_policy: "none"
# sysctl names allowed to be used in security policies, comma-separated
allowed_unsafe_sysctls: "net.ipv4.tcp_keepalive_time,net.ipv4.tcp_keepalive_intvl,net.ipv4.tcp_keepalive_probes,net.ipv4.tcp_syn_retries,net.ipv4.tcp_retries2"
# the maximum amount of memory for EBS CSI controller's sidecars
ebs_csi_controller_sidecar_memory: "80Mi"
ebs_csi_controller_sidecar_cpu: "10m"
# pull images in parallel
serialize_image_pulls: "false"
# rate of image pull in the kubelet, see
# see https://github.com/kubernetes/kubernetes/blob/v1.31.0/staging/src/k8s.io/kubelet/config/v1beta1/types.go#L200-L212
#
# registryPullQPS is the limit of registry pulls per second.
# The value must not be a negative number.
# Setting it to 0 means no limit.
# Default: 5
kubelet_registry_pull_qps : "20"
# registryBurst is the maximum size of bursty pulls, temporarily allows
# pulls to burst to this number, while still not exceeding registryPullQPS.
# The value must not be a negative number.
# Only used if registryPullQPS is greater than 0.
# Default: 10
kubelet_registry_burst: "40"
# Version of the scheduler or controller-manager used by the master nodes.
# Supported values:
# - upstream: official Kubernetes version
# - zalando: internal Zalando build with our custom patches
kubernetes_scheduler_image: "zalando"
kubernetes_controller_manager_image: "upstream"
# when set to true, service account tokens can be used from outside the cluster
allow_external_service_accounts: "false"
# issue tokens with a long expiration time in order to detect applications that don't refresh tokens.
rotate_service_account_tokens_extended_expiration: "true"
# Comma separated list of dimensions to include in the Prometheus metrics
# exposed by audittrail-adapter.
# Adding more dimensions can help detect which clients are calling the
# Kubernetes API. Examples:
# * Detect clients not rotating service tokens: `authentication_stale_token` metric.
# * Detect clients with high load on the API Server for certain calls.
# * Detect client calling old api_version/api_groups for resources.
#
# Adding more dimensions has the negative effect that it produces more
# Prometheus data, so it's not intended to be enabled ALL THE TIME, but can be
# enabled when needed to answer one of the above questions or similar.
#
# The possible dimensions are:
# authorization_decision,authentication_stale_token,user,user_agent,verb,code,resource,api_group,api_version
auditlog_metric_dimensions: "authorization_decision"
# enable auditlogging of read access to identify service accounts reading from
# the api.
auditlog_read_access: "false"
# allow ssh access for internal VPC IPs only
ssh_vpc_only: "true"
# configure custom dns zone
custom_dns_zone: "" # zone name e.g. example.org
custom_dns_zone_nameservers: "" # space separated list of nameserver IP addresses
# prefix prepended to ownership TXT records for external-dns
external_dns_ownership_prefix: "_external-dns."
# domains that should be included by ExternalDNS ("" includes all hosted zones in the account. Separate multiple domains with a comma)
external_dns_domain_filter: ""
# domains that should be ignored by ExternalDNS
external_dns_excluded_domains: cluster.local
# synchronization policy between Kubernetes and AWS Route53 (default: sync, options: sync, upsert-only, create-only)
external_dns_policy: sync
# the duration for how long to cache the list of hosted zones in memory
external_dns_zones_cache_duration: "1h"
# resource configuration
external_dns_mem: "4Gi"
# select which cache to use for Cluster DNS: unbound or dnsmasq.
dns_cache: "dnsmasq"
expirimental_dns_unbound_liveness_probe: "true"
# DNS container resources
dns_dnsmasq_cpu: "100m"
dns_dnsmasq_mem: "50Mi"
dns_dnsmasq_sidecar_cpu: "10m"
dns_dnsmasq_sidecar_mem: "45Mi"
dns_unbound_cpu: "100m"
dns_unbound_mem: "50Mi"
dns_unbound_telemetry_cpu: "10m"
dns_unbound_telemetry_mem: "45Mi"
dns_coredns_cpu: "50m"
dns_coredns_mem: "100Mi"
# special roles for test/pet clusters
{{if eq .Cluster.Environment "e2e"}}
collaborator_administrator_access: "true"
{{else}}
collaborator_administrator_access: "false"
{{end}}
node_auth_rate_limit: "5.0"
# enable legacy serviceaccounts for smooth RBAC migration
enable_operator_sa: "false"
enable_default_sa: "false"
# virtual memory configuration
vm_dirty_background_bytes: "67108864"
vm_dirty_bytes: "134217728"
# Option to Enable FeatureGate TopologyAwareHints
enable_topology_aware_hints: "false"
# Enable FeatureGate HPAScaleToZero
enable_hpa_scale_to_zero: "true"
# Enable FeatureGate MaxUnavailableStatefulSet
max_unavailable_statefulset_enabled: "false"
# enable encryption of secrets in etcd
# this flag can be switched between true and false
# to ensure all secrets are encrypted/decrypted all secrets need to be rewritten after masters have been rolled
enable_encryption: "true"
# enable non-default profiles for the default scheduler
enable_scheduler_profiles: "true"
# default ttl for kube janitor for resources build from PRs
kube_janitor_default_pr_ttl: "1w" # 1 week
# enable cleanup of pr resources other than namespaces
kube_janitor_cleanup_pr_resources: "true"
# opt-in deletion of unused PVCs
kube_janitor_default_unused_pvc_ttl: "forever"
# deletes all resources in the cluster that rely on a vpc
# necessary to change the VPC subnet of a cluster
delete_vpc_resources: "false"
# replacement strategy used for default on-demand worker pool
on_demand_worker_replacement_strategy: none
# SpotAllocationStrategy for pools
spot_allocation_strategy: "capacity-optimized"
# Stackset controller
stackset_controller_sync_interval: "10s"
stackset_controller_mem_min: "120Mi"
stackset_controller_mem_max: "4Gi"
# EBS settings for the root volume
ebs_master_root_volume_size: "50"
ebs_root_volume_size: "100"
ebs_root_volume_delete_on_termination: "true"
# Priority class used for critical system pods
system_priority_class: "cluster-critical-nonpreempting"
# configuration for the PDB controller
{{if eq .Cluster.Environment "test" }}
pdb_controller_non_ready_ttl: "1h"
{{else}}
pdb_controller_non_ready_ttl: ""
{{end}}
pdb_controller_max_unavailable: "1%"
# Log Kubernetes events to Scalyr
kubernetes_event_logger_enabled: "true"
event_logger_mem_min: "100Mi"
event_logger_cpu_min: "10m"
# enable/disable routegroup support for stackset
stackset_routegroup_support_enabled: "true"
# enable/disable inline configmap support for stackset
stackset_inline_configmap_support_enabled: "false"
# enable/disable platformCredentialsSet support for stackset
stackset_pcs_support_enabled: "true"
# Enable/Disable profiling for Kubernetes components
enable_control_plane_profiling: "false"
okta_auth_issuer_url: ""
okta_auth_client_id: "kubernetes.cluster.{{.Cluster.Alias}}"
# Deploy
# This section contains config items to enable and disable the the