daemon, fqdn: Add flag to control FQDN regex LRU size #19383

christarazi · 2022-04-09T00:16:26Z

Advanced users can configure the LRU size for the cache holding the
compiled regex expressions of FQDN match{Pattern,Name}. This is useful
if users are experiencing high memory usage spikes with many FQDN
policies that have repeated matchPattern or matchName across many
different policies.

Please see second commit msg for the performance benchmark results.

Signed-off-by: Chris Tarazi chris@isovalent.com

Related: #17224

christarazi · 2022-04-11T21:42:47Z

pkg/fqdn/dnsproxy/proxy_test.go

+	b.Skip()
+
+	bb, err := ioutil.ReadFile("testdata/cnps-large.yaml")
+	if err != nil {
+		b.Fatal(err)
+	}


For now this benchmark will be skipped until we have a proper fixture for CNPs checked in.

christarazi · 2022-04-11T22:20:13Z

/test

Job 'Cilium-PR-K8s-1.21-kernel-5.4' failed:

Click to show.

Test Name

K8sDatapathConfig AutoDirectNodeRoutes Check connectivity with sockops and direct routing

Failure Output

FAIL: Error creating resource /home/jenkins/workspace/Cilium-PR-K8s-1.21-kernel-5.4/src/github.com/cilium/cilium/test/k8s/manifests/l3-policy-demo.yaml: Cannot retrieve cilium pod cilium-w9p78 policy revision: cannot get revision from json output 'Defaulted container "cilium-agent" out of: cilium-agent, mount-cgroup (init), clean-cilium-state (init)

If it is a flake and a GitHub issue doesn't already exist to track it, comment /mlh new-flake Cilium-PR-K8s-1.21-kernel-5.4 so I can create one.

Edit:

net-next failed to provision
5.4 hit CI: K8sDatapathConfig AutoDirectNodeRoutes Check connectivity with sockops and direct routing #16852

christarazi · 2022-04-11T23:24:21Z

/test-1.23-net-next

Job 'Cilium-PR-K8s-GKE' failed:

Click to show.

Test Name

K8sDatapathConfig Host firewall With native routing

Failure Output

FAIL: Managed to reach 10.48.2.1:69 from testclient-host-mp9p4

If it is a flake and a GitHub issue doesn't already exist to track it, comment /mlh new-flake Cilium-PR-K8s-GKE so I can create one.

Edit: hit #16159

christarazi · 2022-04-27T19:45:03Z

/test-1.21-5.4

christarazi · 2022-04-28T18:26:48Z

Will push short to resolve conflict. Noting that CI has passed besides known flakes.

Advanced users can configure the LRU size for the cache holding the compiled regex expressions of FQDN match{Pattern,Name}. This is useful if users are experiencing high memory usage spikes with many FQDN policies that have repeated matchPattern or matchName across many different policies. Signed-off-by: Chris Tarazi <chris@isovalent.com>

When comparing efficieny of increasing the LRU size from 128 to 1024 with ~22k CNPs, we see the following results: ``` \# LRU size 128. $ go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x -memprofile memprofile.out ./pkg/fqdn/dnsproxy > old.txt \# LRU size 1024. $ go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x -memprofile memprofile.out ./pkg/fqdn/dnsproxy > new.txt $ benchcmp old.txt new.txt benchcmp is deprecated in favor of benchstat: https://pkg.go.dev/golang.org/x/perf/cmd/benchstat benchmark old ns/op new ns/op delta Benchmark_perEPAllow_setPortRulesForID_large-8 3954101340 3010934555 -23.85% benchmark old allocs new allocs delta Benchmark_perEPAllow_setPortRulesForID_large-8 26480632 24167742 -8.73% benchmark old bytes new bytes delta Benchmark_perEPAllow_setPortRulesForID_large-8 2899811832 1824062992 -37.10% ``` Here's the raw test run with LRU size at 128: ``` Before (N=1) Alloc = 31 MiB HeapInuse = 45 MiB Sys = 1260 MiB NumGC = 15 After (N=1) Alloc = 445 MiB HeapInuse = 459 MiB Sys = 1260 MiB NumGC = 40 ``` Here's the raw test run with LRU size at 1024: ``` Before (N=1) Alloc = 31 MiB HeapInuse = 48 MiB Sys = 1177 MiB NumGC = 17 After (N=1) Alloc = 78 MiB HeapInuse = 93 MiB Sys = 1177 MiB NumGC = 53 ``` We can see that it's saving ~300MB. Furthermore, if we compare the memprofiles from the benchmark run via ``` go tool pprof -http :8080 -diff_base memprofile.out memprofile.1024.out ``` we see an ~800MB reduction in the regex compilation. Signed-off-by: Chris Tarazi <chris@isovalent.com>

Following the previous commit's benchmark result, let's update the LRU default size to be 1024, given that it only results in a few 10's of MBs increase when the cache nears full. Signed-off-by: Chris Tarazi <chris@isovalent.com>

christarazi · 2022-04-29T00:18:15Z

Travis passed as a sanity check for the rebase. Marking ready to merge given #19383 (comment).

Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: cilium#19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com>

Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: #19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com>

[ upstream commit 6c0cc03 ] Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: cilium#19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com> Signed-off-by: Louis DeLosSantos <louis.delos@isovalent.com>

[ upstream commit 6c0cc03 ] Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: #19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com> Signed-off-by: Louis DeLosSantos <louis.delos@isovalent.com>

[ upstream commit 6c0cc03 ] Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: cilium#19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com> Signed-off-by: Louis DeLosSantos <louis.delos@isovalent.com> Signed-off-by: Michi Mutsuzaki <michi@isovalent.com>

[ upstream commit 6c0cc03 ] Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: cilium#19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com> Signed-off-by: Nicolas Busseneau <nicolas@isovalent.com>

Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: cilium#19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com>

[ upstream commit 6c0cc03 ] Previously, the printed memory values depended greatly on both the size of the loaded CNPs in memory, as well as when the gc ran for the last time. That favored solutions that allocated a lot of memory in the start of the test (but still inside the actual testing section), while solution allocating a lot of memory in the end, even if they removed all the referneces to it, were punished due the fact that we printed the memory stats before forcing a gc. For heap, the diff between the "Before Setup" and "After Test" should be inspected, since "Before Test" has all the CNPs in memory. When we hit "After test", those CNPs have been garbage collected. New output: root@10fc3f693c5d:/code# go test -tags privileged_tests -v -run '^$' -bench Benchmark_perEPAllow_setPortRulesForID_large -benchmem -benchtime 1x ./pkg/fqdn/dnsproxy goos: linux goarch: amd64 pkg: github.com/cilium/cilium/pkg/fqdn/dnsproxy cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz Benchmark_perEPAllow_setPortRulesForID_large Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 5 MiB Sys = 20 MiB NumGC = 6 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 50 MiB Sys = 987 MiB NumGC = 16 After Test (N=1,EPs=20,cache=128) Alloc = 341 MiB HeapInuse = 374 MiB Sys = 987 MiB NumGC = 38 Before Setup (N=1,EPs=20,cache=128) Alloc = 3 MiB HeapInuse = 6 MiB Sys = 987 MiB NumGC = 40 Before Test (N=1,EPs=20,cache=128) Alloc = 32 MiB HeapInuse = 56 MiB Sys = 1020 MiB NumGC = 49 After Test (N=1,EPs=20,cache=128) Alloc = 344 MiB HeapInuse = 378 MiB Sys = 1020 MiB NumGC = 71 Benchmark_perEPAllow_setPortRulesForID_large-7 1 4301736800 ns/op 3047525016 B/op 26526920 allocs/op PASS ok github.com/cilium/cilium/pkg/fqdn/dnsproxy 20.796s Fixes: #19383 Fixes: 38c0036 ("dnsproxy: Add benchmark for large FQDN-based CNPs") Signed-off-by: Odin Ugedal <ougedal@palantir.com> Signed-off-by: Nicolas Busseneau <nicolas@isovalent.com>

christarazi requested a review from a team April 9, 2022 00:16

christarazi requested a review from a team as a code owner April 9, 2022 00:16

christarazi requested review from jibi and michi-covalent April 9, 2022 00:16

maintainer-s-little-helper bot added dont-merge/needs-release-note-label The author needs to describe the release impact of these changes. and removed dont-merge/needs-release-note-label The author needs to describe the release impact of these changes. labels Apr 9, 2022

christarazi requested a review from aanm April 9, 2022 00:17

aanm approved these changes Apr 9, 2022

View reviewed changes

christarazi commented Apr 11, 2022

View reviewed changes

aanm approved these changes Apr 11, 2022

View reviewed changes

christarazi added the kind/performance There is a performance impact of this. label Apr 11, 2022

christarazi added the needs-backport/1.11 label Apr 21, 2022

jibi approved these changes Apr 26, 2022

View reviewed changes

christarazi added 3 commits April 28, 2022 11:33

dameon: Change the default FQDN regex LRU to be 1024

676e87c

Following the previous commit's benchmark result, let's update the LRU default size to be 1024, given that it only results in a few 10's of MBs increase when the cache nears full. Signed-off-by: Chris Tarazi <chris@isovalent.com>

christarazi force-pushed the pr/christarazi/fqdn-regex-lru-flag branch from c8cb648 to 676e87c Compare April 28, 2022 18:33

christarazi mentioned this pull request Apr 28, 2022

Use FQDN regex LRU everywhere #19632

Merged

christarazi added the ready-to-merge This PR has passed all tests and received consensus from code owners to merge. label Apr 29, 2022

maintainer-s-little-helper bot added this to Needs backport from master in 1.11.5 Apr 29, 2022

aditighag merged commit ce9583d into cilium:master Apr 29, 2022

christarazi deleted the pr/christarazi/fqdn-regex-lru-flag branch April 29, 2022 16:30

aditighag mentioned this pull request May 2, 2022

v1.11 backports 2022-05-02 #19671

Merged

aditighag added backport-pending/1.11 and removed needs-backport/1.11 labels May 2, 2022

maintainer-s-little-helper bot moved this from Needs backport from master to Backport pending to v1.10 in 1.11.5 May 2, 2022

aditighag added backport-done/1.11 The backport for Cilium 1.11.x for this PR is done. and removed backport-pending/1.11 labels May 4, 2022

maintainer-s-little-helper bot moved this from Backport pending to v1.10 to Backport done to v1.11 in 1.11.5 May 4, 2022

aanm mentioned this pull request May 10, 2022

Prepare for release v1.11.5 #19756

Merged

michi-covalent added the needs-backport/1.10 label Jun 6, 2022

joamaki mentioned this pull request Jun 7, 2022

v1.10 backports 2022-06-07 #20100

Merged

joamaki added backport-pending/1.10 and removed needs-backport/1.10 labels Jun 7, 2022

joestringer mentioned this pull request Jun 10, 2022

Prepare for release v1.10.12 #20168

Merged

odinuge mentioned this pull request Jun 28, 2022

dnsproxy: update dnsproxy benchmark memory calculation #20305

Merged

6 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

daemon, fqdn: Add flag to control FQDN regex LRU size #19383

daemon, fqdn: Add flag to control FQDN regex LRU size #19383

christarazi commented Apr 9, 2022 •

edited

christarazi Apr 11, 2022

christarazi commented Apr 11, 2022 •

edited

Test Name

Failure Output

christarazi commented Apr 11, 2022 •

edited

Test Name

Failure Output

christarazi commented Apr 27, 2022

christarazi commented Apr 28, 2022

christarazi commented Apr 29, 2022

daemon, fqdn: Add flag to control FQDN regex LRU size #19383

daemon, fqdn: Add flag to control FQDN regex LRU size #19383

Conversation

christarazi commented Apr 9, 2022 • edited

christarazi Apr 11, 2022

Choose a reason for hiding this comment

christarazi commented Apr 11, 2022 • edited

Test Name

Failure Output

christarazi commented Apr 11, 2022 • edited

Test Name

Failure Output

christarazi commented Apr 27, 2022

christarazi commented Apr 28, 2022

christarazi commented Apr 29, 2022

christarazi commented Apr 9, 2022 •

edited

christarazi commented Apr 11, 2022 •

edited

christarazi commented Apr 11, 2022 •

edited