Eth2 Kubernetes — Node size


1
2
3		Research on the trade offs when choosing an instance type for a Kubernetes cluster for Eth2 Nodes
4
5		Legend:	Blue = Price Optimal
6	Manifold Finance, Inc.		Red = General
7	Telegram	Version:	2022.05.03
8	Discourse Forums
9		Data Source:	learnk8s.io
10
11
12				What is it?	Not all memory and CPU in a Node can be used to run Pods. The resources are partitioned in 4: 1. Memory and CPU reserved to the operating system and system daemons such as SSH 2. Memory and CPU reserved to the Kubelet and Kubernetes agents such as the CRI 3. Memory reserved for the hard eviction threshold 4. Memory and CPU available to Pods The graph show the total memory available for running Pods after subtracting the reserved memory	Memory (in GiB)	GKE	EKS	AKS
13						1	55.00%	50.10%	0.00%
14						2	65.00%	71.20%	32.50%
15						4	70.00%	83.95%	53.75%
16						8	75.00%	90.33%	66.88%
17						16	82.50%	93.17%	78.44%
18						64	91.13%	95.27%	90.11%
19						128	92.56%	97.63%	92.05%
20						192	93.88%	98.42%	93.54%
21						256	94.91%	96.66%	94.65%
22
23				Example	If you have a Kubernetes cluster in GKE with a single Node of 2GB of memory, only 65% of the available memory is used to run Pods. The reamining memory is necessary to run the OS, Kubelet, CRI, CNI, etc.
24
25
26
27				Notes	GKE and AKS reach 90% level of utilisation with instances over 64GB. EKS is 90% efficient starting with 8GB.
28
29
30
31				What is it?	Not all memory and CPU in a Node can be used to run Pods. The resources are partitioned in 4: 1. Memory and CPU reserved to the operating system and system daemons such as SSH 2. Memory and CPU reserved to the Kubelet and Kubernetes agents such as the CRI 3. Memory reserved for the hard eviction threshold 4. Memory and CPU available to Pods The graph show the total memory available for running Pods after subtracting the reserved memory	CPU (in millicores)	GKE	EKS	AKS
32						1	84.00%	84.00%	84.00%
33						2	91.50%	91.50%	90.00%
34						4	95.50%	95.50%	94.00%
35						8	97.63%	97.63%	96.50%
36						16	98.69%	98.69%	97.75%
37						32	99.22%	99.22%	98.38%
38						64	99.48%	99.48%	98.69%
39
40
41
42				Example	If you have a Kubernetes cluster in AKS with a single Node and 2 vCPU 90% of the available CPUs are used to run Pods. The reamining memory is necessary to run the OS, Kubelet, CRI, CNI, etc.
43
44
45
46				Notes	As long as you use node with at least 2 vCPU you should be fine.
47
48
49
50				What is it?	There's a upper limit on the number of Pods that you can run on each Node. Each cloud provider has a different limit. Most of the time the limit is independent of the Node size (e.g. GKE, AKS). There are cases where the number of Pods depends on the Node size (notable: EKS).	Memory (in GiB)	GKE	EKS	AKS
51						1	110	4	250
52						2	110	11	250
53						4	110	17	250
54						8	110	29	250
55						16	110	58	250
56						64	110	234	250
57						128	110	234	250
58						192	110	234	250
59				Notes	The metrics is relevant to measure your blast radius. Assuming that a Node is lost how many Pods are affected?	256	110	737	250
60
61
62
63
64
65
66
67
68
69				What is it?	Nodes have an upper limit on the number of Pods that they can run. Assuming that you run the max number of Pods for that node, how much memory is available to each Pod? This metric divides the available Node memory by the max number of Pods for that instance type.	Memory (in GiB)	GKE	EKS	AKS
70						1	0.01	0.13	0.00
71						2	0.01	0.13	0.00
72						4	0.03	0.20	0.01
73						8	0.05	0.25	0.02
74						16	0.12	0.26	0.05
75						64	0.53	0.26	0.23
76				Example	If you have Kubernetes cluster in GKE with a single Node of 128GB of memory, you can run up to 110 Pods and each of them can use 1.08GB of memory.	128	1.08	0.53	0.47
77						192	1.64	0.81	0.72
78						256	2.21	0.34	0.97
79
80				Notes	It's not possible to run small workloads (less than 1GB of memory) efficiently on GKE when the node size is greater than 128GB of memory. EKS has a peak at 192GB of memory. That's where there are the most Pod with the larger memory available to them (234 Pods with 810MiB of memory each).
81
82
83
84
85
86
87
88				What is it?	If all my Pods are using 1GB of memory, what instance type I should use to maximise the memory available? The charts presents 5 scenarios: what if all the Pods in the Node have limits of 1, 2, 4, 8 or 16 GiB. The chart shows how utilised is the node.	Pod memory limit	1GiB	2GiB	4GiB	8GiB	16GiB	64GiB	128 GiB	192 GiB	256 GiB
89						0.5GiB	50.00%	50.00%	62.50%	75.00%	81.25%	85.94%	42.97%	28.65%	21.48%
90						1 GiB	0.00%	50.00%	50.00%	75.00%	81.25%	90.63%	85.94%	57.29%	42.97%
91						2 GiB	0.00%	0.00%	50.00%	75.00%	75.00%	90.63%	92.19%	93.75%	85.94%
92						4 GiB	0.00%	0.00%	0.00%	50.00%	75.00%	87.50%	90.63%	93.75%	93.75%
93						8 GiB	0.00%	0.00%	0.00%	0.00%	50.00%	87.50%	87.50%	91.67%	93.75%
94						16 GiB	0.00%	0.00%	0.00%	0.00%	0.00%	75.00%	87.50%	91.67%	93.75%
95
96				Example	When all Pods in your cluster are 1GB, the best node that can allocate the most number of Pods is a Node with 64GB of memory. Values before the peak means that the node is underutilised (there's still space, but not enough to run a Pod). Values after the peak means that you reached the limit of Pods on that and you can't schedule more Pods on that node.
97
98
99
100
101
102
103
104
105				Notes	It's clear that the best Node for Pods that average 1GB of memory is a 64GB Node. If the Pod memory limit increases in average to 2GB, a 192GB memory instance is the more efficient.
106
107
108
109
110
111
112				What is it?	If all my Pods are using 1GB of memory, what instance type I should use to maximise the memory available? The charts presents 5 scenarios: what if all the Pods in the Node have limits of 1, 2, 4, 8 or 16 GiB. The chart shows how utilised is the node.	Pod memory limit	1GiB	2GiB	4GiB	8GiB	16GiB	64GiB	128 GiB	192 GiB	256 GiB
113						0.5GiB	50.00%	50.00%	75.00%	87.50%	90.63%	94.53%	91.41%	60.94%	96.48%
114						1 GiB	0.00%	50.00%	75.00%	87.50%	87.50%	93.75%	96.88%	97.92%	96.48%
115						2 GiB	0.00%	0.00%	50.00%	75.00%	87.50%	93.75%	96.88%	97.92%	96.09%
116						4 GiB	0.00%	0.00%	0.00%	50.00%	75.00%	93.75%	96.88%	97.92%	95.31%
117						8 GiB	0.00%	0.00%	0.00%	0.00%	50.00%	87.50%	93.75%	95.83%	93.75%
118						16 GiB	0.00%	0.00%	0.00%	0.00%	0.00%	75.00%	87.50%	91.67%	93.75%
119
120				Example	When all Pods in your cluster are 1GB, the best node that can allocate the most number of Pods is a Node with 192GB of memory. Values before the peak means that the node is underutilised (there's still space, but not enough to run a Pod).
121
122
123
124
125
126
127				Notes	Pay attention to local inefficiencies due to the limits on how many Pods can be deployed on a Node.
128
129
130
131
132				What is it?	If all my Pods are using 1GB of memory, what instance type I should use to maximise the memory available? The charts presents 5 scenarios: what if all the Pods in the Node have limits of 1, 2, 4, 8 or 16 GiB. The chart shows how utilised is the node.	Pod memory limit	1GiB	2GiB	4GiB	8GiB	16GiB	64GiB	128 GiB	192 GiB	256 GiB
133						0.5GiB	0.00%	25.00%	50.00%	62.50%	78.13%	89.84%	91.80%	65.10%	48.83%
134						1 GiB	0.00%	25.00%	50.00%	62.50%	78.13%	89.84%	91.80%	65.10%	48.83%
135						2 GiB	0.00%	0.00%	50.00%	62.50%	75.00%	89.06%	91.41%	93.23%	94.53%
136						4 GiB	0.00%	0.00%	50.00%	50.00%	75.00%	87.50%	90.63%	92.71%	94.53%
137						8 GiB	0.00%	0.00%	0.00%	50.00%	75.00%	87.50%	90.63%	91.67%	93.75%
138						16 GiB	0.00%	0.00%	0.00%	0.00%	50.00%	87.50%	87.50%	91.67%	93.75%
139
140				Example	When all Pods in your cluster are 1GB, the best node that can allocate the most number of Pods is a Node with 128GB of memory. Values before the peak means that the node is underutilised (there's still space, but not enough to run a Pod). Values after the peak means that you reached the limit of Pods on that and you can't schedule more Pods on that node.
141
142
143
144
145
146
147
148
149
150
151				What is it?	Nodes have an upper limit on the number of Pods that they can run. Assuming that you run the max number of Pods for that node, how much CPU is available to each Pod? This metric divides the available Node cores by the max number of Pods for that instance type.	CPU cores	GKE	EKS	AKS
152						1	7.64	210.00	3.36
153						2	16.64	63.10	7.20
154						4	34.73	65.86	15.04
155						8	71.00	134.66	30.88
156						16	143.55	67.48	62.56
157						32	288.64	135.68	125.92
158						64	578.82	86.39	252.64
159				Example	GKE can have at most 110 pods per node. If the Node has 16 cores, you can split the available millicores evenly to all Pods. Each Pod receives a quota of 143.55 millicores.
160
161
162
163
164
165
166				Notes	Larger nodes in GKE and AKS lead to Pod with more CPU available (due to the cap on the number of Pods)
167
168
169
170
171				What is it?	Blockchain Node Calculator	Blockchain Nodes	GKE	EKS	AKS
172						1
173						2
174						4
175						8
176						16
177						32
178						64
179				Example
180
181
182
183
184
185
186				Notes
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


1	Concensus Types	IBFT2	CLIQUE	AURA	RAFT
2		Besu	Besu	N/A	N/A
3	Security
4	Liveness assumption (e.g. watch-towers)	Yes	Bonded	N/A	N/A
5	The mass exit assumption	No	No	N/A	N/A
6	Quorum of validators can freeze funds	Yes	Yes	N/A	N/A
7	Quorum of validators can confiscate funds	Yes	Yes	N/A	N/A
8	Vulnerability to hot-wallet key exploits	Moderate	Moderate	N/A	N/A
9	Vulnerability to crypto-economic attacks	Moderate	Moderate	N/A	N/A
10	Cryptographic primitives	Standard	Standard	N/A	N/A
11				N/A	N/A
12	Performance / economics			N/A	N/A
13	Max throughput on ETH 1.0	10k+ TPS	10k+ TPS	N/A	N/A
14	Max throughput on ETH 2.0*	1..∞ TPS ²	1..∞ TPS ²	N/A	N/A
15	Capital-efficient	3/4 Nodes Req.	51% of Nodes	N/A	N/A
16	Separate onchain tx to open new account	N/A	N/A	N/A	N/A
17	Cost of tx	Very low	Low	N/A	N/A
18	Liveness	Does Not Favor	Favors	N/A	N/A
19	Finality	Favors	Does Not Favor	N/A	N/A
20	Edge Cases	Yes	Yes	N/A	N/A
21
22	Usability
23	Withdrawal time	1 confirm.	x confirm.	N/A	N/A
24	Time to subjective finality	Instant	[ref]	N/A	N/A
25	Client-side verification of subjective finality	Yes	N/A (trusted)	N/A	N/A
26	Instant tx confirmations	Full	Bonded	N/A	N/A
27
28
29	* TBD, Estimation


1		Millicores reserved for the kubelet and kubernetes agents
2	CPU cores	EKS
3		Kubelet	OS	Total reserved	Total reserved %	Available	Available %
4	1	60	100	160	16.00%	840	84.00%
5	2	70	100	170	8.50%	1830	91.50%
6	4	80	100	180	4.50%	3820	95.50%
7	8	90	100	190	2.38%	7810	97.63%
8	16	110	100	210	1.31%	15790	98.69%
9	32	150	100	250	0.78%	31750	99.22%
10	64	230	100	330	0.52%	63670	99.48%
11
12
13	Memory
14	EKS	Instance type	Max Pods per node	Kubelet	OS	Hard eviction	Total Reserved	Total Reserved %	Available for workloads	Available for workloads %	Max density
15	1	t2.micro	4	0.299	0.1	0.1	0.499	49.90%	0.501	50.10%	0.12525
16	2	t2.small	11	0.376	0.1	0.1	0.576	28.80%	1.424	71.20%	0.1294545455
17	4	t2.medium	17	0.442	0.1	0.1	0.642	16.05%	3.358	83.95%	0.1975294118
18	8	m5.large	29	0.574	0.1	0.1	0.774	9.68%	7.226	90.33%	0.2491724138
19	16	m5.xlarge	58	0.893	0.1	0.1	1.093	6.83%	14.907	93.17%	0.2570172414
20	64	m5.4xlarge	234	2.829	0.1	0.1	3.029	4.73%	60.971	95.27%	0.2605598291
21	128	m5.8xlarge	234	2.829	0.1	0.1	3.029	2.37%	124.971	97.63%	0.5340641026
22	192	m5.12xlarge	234	2.829	0.1	0.1	3.029	1.58%	188.971	98.42%	0.8075683761
23	256	m5.16xlarge	737	8.362	0.1	0.1	8.562	3.34%	247.438	96.66%	0.3357367707
24
25	Pods per Node (with memory limits)
26	Instance memory		1	2	4	8	16	64	128	192	256
27	Instance memory available for pods		0.501	1.424	3.358	7.226	14.907	60.971	124.971	188.971	247.438
28	Pod memory size	0.5	1	2	6	14	29	121	249	377	494
29		1	0	1	3	7	14	60	124	188	247
30		2	0	0	1	3	7	30	62	94	123
31		4	0	0	0	1	3	15	31	47	61
32		8	0	0	0	0	1	7	15	23	30
33		16	0	0	0	0	0	3	7	11	15
34
35	Pods per Node (with memory limits) CAPPED
36	Instance memory		1	2	4	8	16	64	128	192	256
37	Max pods per node		4	11	17	29	58	234	234	234	737
38	Pod memory size	0.5	1	2	6	14	29	121	234	234	494
39		1	0	1	3	7	14	60	124	188	247
40		2	0	0	1	3	7	30	62	94	123
41		4	0	0	0	1	3	15	31	47	61
42		8	0	0	0	0	1	7	15	23	30
43		16	0	0	0	0	0	3	7	11	15
44
45	Pods per Node (with memory limits) Efficiency
46	Instance memory		1	2	4	8	16	64	128	192	256
47	Pod memory size	0.5	50.00%	50.00%	75.00%	87.50%	90.63%	94.53%	91.41%	60.94%	96.48%
48		1	0.00%	50.00%	75.00%	87.50%	87.50%	93.75%	96.88%	97.92%	96.48%
49		2	0.00%	0.00%	50.00%	75.00%	87.50%	93.75%	96.88%	97.92%	96.09%
50		4	0.00%	0.00%	0.00%	50.00%	75.00%	93.75%	96.88%	97.92%	95.31%
51		8	0.00%	0.00%	0.00%	0.00%	50.00%	87.50%	93.75%	95.83%	93.75%
52		16	0.00%	0.00%	0.00%	0.00%	0.00%	75.00%	87.50%	91.67%	93.75%
53
54	CPU cores	Instance type	Max Pods per node	CPU available	CPU quota per Pod (in millicores)
55	1	t2.micro	4	840	210
56	2	m5.large	29	1830	63.10344828
57	4	m5.xlarge	58	3820	65.86206897
58	8	m5.2xlarge	58	7810	134.6551724
59	16	m5.4xlarge	234	15790	67.47863248
60	32	m5.8xlarge	234	31750	135.6837607
61	64	m5.16xlarge	737	63670	86.39077341