@@ -1273,6 +1273,7 @@ No.
1273
1273
1274
1274
# ##### Will enabling / using this feature result in increasing time taken by any operations covered by existing SLIs/SLOs?
1275
1275
1276
+
1276
1277
<!--
1277
1278
Look at the [existing SLIs/SLOs].
1278
1279
@@ -1281,8 +1282,22 @@ Think about adding additional work or introducing new steps in between
1281
1282
1282
1283
[existing SLIs/SLOs] : https://git.k8s.io/community/sig-scalability/slos/slos.md#kubernetes-slisslos
1283
1284
-->
1284
- Performance of any writes to a custom resource may be impacted. Benchmarks must
1285
- be taken to measure magniture.
1285
+ The benchmark has been written in https://github.com/kubernetes/kubernetes/pull/121405. The difference in time is minimal.
1286
+ ` ` ` shell
1287
+ ~ go test -bench=. ./staging/src/k8s.io/apiextensions-apiserver/test/integration/ -run=BenchmarkRatcheting
1288
+ goos: linux
1289
+ goarch: amd64
1290
+ pkg: k8s.io/apiextensions-apiserver/test/integration
1291
+ cpu: AMD EPYC 7B13
1292
+ BenchmarkRatcheting/RatchetingEnabled/ValidXValid-128 1 1250442700 ns/op
1293
+ BenchmarkRatcheting/RatchetingEnabled/ValidXInvalid-128 2 582064833 ns/op
1294
+ BenchmarkRatcheting/RatchetingEnabled/InvalidXInvalid-128 7 161029972 ns/op
1295
+ BenchmarkRatcheting/RatchetingDisabled/ValidXValid-128 1 1203940147 ns/op
1296
+ BenchmarkRatcheting/RatchetingDisabled/ValidXInvalid-128 2 574053662 ns/op
1297
+ BenchmarkRatcheting/RatchetingDisabled/InvalidXInvalid-128 7 153500634 ns/op
1298
+ PASS
1299
+ ok k8s.io/apiextensions-apiserver/test/integration 8.842s
1300
+ ` ` `
1286
1301
1287
1302
# ##### Will enabling / using this feature result in non-negligible increase of resource usage (CPU, RAM, disk, IO, ...) in any components?
1288
1303
@@ -1295,7 +1310,22 @@ This through this both in small and large cases, again with respect to the
1295
1310
1296
1311
[supported limits] : https://git.k8s.io/community//sig-scalability/configs-and-limits/thresholds.md
1297
1312
-->
1298
- Should have benchmarks before beta, but expecting to see some measureable impact to writes only to CRDs.
1313
+ The benchmark has been written in https://github.com/kubernetes/kubernetes/pull/121405. The difference in resource usage is minimal.
1314
+ ` ` ` shell
1315
+ go test -bench=. ./staging/src/k8s.io/apiextensions-apiserver/test/integration/ -run=BenchmarkRatcheting -benchmem
1316
+ goos: linux
1317
+ goarch: amd64
1318
+ pkg: k8s.io/apiextensions-apiserver/test/integration
1319
+ cpu: AMD EPYC 7B13
1320
+ BenchmarkRatcheting/RatchetingEnabled/ValidXValid-128 1 1290557385 ns/op 525375344 B/op 8876573 allocs/op
1321
+ BenchmarkRatcheting/RatchetingEnabled/ValidXInvalid-128 2 593485537 ns/op 248714944 B/op 4217124 allocs/op
1322
+ BenchmarkRatcheting/RatchetingEnabled/InvalidXInvalid-128 7 163805382 ns/op 68990193 B/op 1179765 allocs/op
1323
+ BenchmarkRatcheting/RatchetingDisabled/ValidXValid-128 1 1227604853 ns/op 511288592 B/op 8714209 allocs/op
1324
+ BenchmarkRatcheting/RatchetingDisabled/ValidXInvalid-128 2 587069751 ns/op 242705312 B/op 4146673 allocs/op
1325
+ BenchmarkRatcheting/RatchetingDisabled/InvalidXInvalid-128 7 154602244 ns/op 65377312 B/op 1136079 allocs/op
1326
+ PASS
1327
+ ok k8s.io/apiextensions-apiserver/test/integration 10.018s
1328
+ ` ` `
1299
1329
1300
1330
# ##### Can enabling / using this feature result in resource exhaustion of some node resources (PIDs, sockets, inodes, etc.)?
1301
1331
0 commit comments