Add extra addresses for listeners configuration to support dual stack in Istio #40634

zhlsunshine · 2022-08-24T13:58:31Z

Please provide a description of this PR:
Add extra addresses for listeners configuration to support dual stack in Istio, this PR is part of issue#40394

This PR just focus on normal k8s service in service mesh, and add the additional addresses for specified listener of service. There fore, it does not include below:

No code change for service entry
No code change for statefulsets/headless services with TCP ports, and empty service address field (There is another PR for this: wip: experimental with single listener for headless #40409)
No code change for gateway listener configuration

Below snapshots are test results and the attach contains the whole configdump file exclude eds via curl localhost:15000/config_dump

Inbound listener additional addresses:

Outbound listener additional addresses:

Application/Service listener additional addresses:

Config dump:
extra_addresses_configdump.log

Note: I did not add any test case and I will add necessary test case after all the business code changes are available for reviewers.

… in Istio

jacob-delgado · 2022-08-24T23:08:12Z

cc @adams-shaun can you help us review this? Do you have any ideas for how we can add tests to this?

howardjohn

Conceptually looks good. I think the way the code is structured right now makes it very error prone for use, I think we need to work on cleaning it up a bit.

pilot/pkg/networking/core/v1alpha3/listener.go

pilot/pkg/networking/core/v1alpha3/listener_builder.go

pilot/pkg/networking/core/v1alpha3/listener_inbound.go

howardjohn · 2022-08-25T14:14:29Z

pilot/pkg/networking/core/v1alpha3/listener_inbound.go


 	// Build the "virtual" inbound listener. This will capture all inbound redirected traffic and contains:
 	// * Passthrough filter chains, matching all unmatched traffic. There are a few of these to handle all cases
 	// * Service filter chains. These will either be for each Port exposed by a Service OR Sidecar.Ingress configuration.
 	allChains := buildInboundPassthroughChains(lb)
 	allChains = append(allChains, chains...)
-	return lb.buildInboundListener(model.VirtualInboundListenerName, util.BuildAddress(actualWildcard, ProxyInboundListenPort), false, allChains)
+	l := lb.buildInboundListener(model.VirtualInboundListenerName, util.BuildAddress(actualWildcard, ProxyInboundListenPort), false, allChains)


should we do this in buildInboundListener?

Hi @howardjohn, I think it's better to separate the default service and extra service address, because 1. no change for buildInboundListener ； 2. no impact for single stack logic; 3. Extra service address should be added separately, because it's not just for dual stack, it may be used for other in future. Any idea?

pilot/pkg/networking/core/v1alpha3/listener.go

howardjohn · 2022-08-25T14:16:49Z

pilot/pkg/networking/core/v1alpha3/listener.go

+				extraAddress := &listener.AdditionalAddress{
+					Address: util.BuildAddress(exbd, uint32(opts.port.Port)),
+				}
+				res.AdditionalAddresses = append(res.AdditionalAddresses, extraAddress)


Do we need a check for Envoy version, this field is new? Or its already in 1.15 and this PR will only be in 1.16?

Hi @howardjohn, yes, AdditionalAddresses is a new field. And yes, but how about to check the Istio version?
BTW, when I execute cmd: make build it's okay with the new field of AdditionalAddresses, but when I execute cmd make test, there are errors message like below:

{Failed listener_builder_test.go:79: listener 127.0.0.1_15007 is invalid: invalid Listener.AdditionalAddresses[0]: embedded message failed validation | caused by: invalid AdditionalAddress.Address: embedded message failed validation | caused by: invalid Address.SocketAddress: embedded message failed validation | caused by: invalid SocketAddress.Address: value length must be at least 1 runes}

So I think make build or make test have different dependency libs version for github.com/envoyproxy/go-control-plane/envoy/config/listener/v3 Any idea?

Added a new function named IsIstioVersionGE116 to make sure that the field of AdditionalAddresses can be added if istioctl version is greater or equal 1.16.

You need go.mod updated to use envoyproxy/go-control-plane@2cb7986

Try running `go get github.com/envoyproxy/go@2cb7986b23ccfde4a77c26ebf3bfbfa71126fafd for your PR.

I have a PR to update the to a version of go.mod, but it's failing due to other reasons.

It appears that updating the api is also blocked on envoyproxy/envoy#22804 as definitions were removed that Istio relies on for other features. Maybe you can get Alex to help you Steve to get it pushed through Envoy. This is blocking Istio as well.

Hi @jacob-delgado, thanks, this error has been fixed!

howardjohn · 2022-08-25T14:17:37Z

pilot/pkg/networking/core/v1alpha3/listener_inbound.go

+					// if it's dual stack environment
+					if bindAddress != "" && exBindAddresses != "" {
+						cc.bind = bindAddress
+						cc.exBind = []string{exBindAddresses}


IIRC before in dual stack we made 2 listeners. Now are we creating 2 listeners with 2 addresses each? Or did e remove the 2 listeners somewhere I missed?

Hi @howardjohn, according to original solution, there are 2 listeners for a service with both IPv4 and IPv6 addresses. For current solution, there is only 1 listener and there are address and additional addresses (new field) for it. The former is for the first address of the service, and the latter is for the rest addresses of the service.

hzxuzhonghu · 2022-08-30T12:12:33Z

cc @hzxuzhonghu @ramaraochavali

zhlsunshine · 2022-08-30T15:23:45Z

Hi @howardjohn, @hzxuzhonghu and @ramaraochavali , I added the unit test and fixed the issues. And here is the new configdump file
new_extra_addresses_configdump.log

howardjohn · 2022-09-02T16:45:53Z

pilot/pkg/networking/core/v1alpha3/listener.go

-			bind := egressListener.IstioListener.Bind
+			var bind string
+			var extraBind []string
+			bind = egressListener.IstioListener.Bind


Suggested change

bind = egressListener.IstioListener.Bind

bind := egressListener.IstioListener.Bind

okay, done!

howardjohn · 2022-09-02T16:47:03Z

pilot/pkg/networking/core/v1alpha3/listener.go

+					// and make sure that it's necessary to add the extra addresses, such as it's a dual stack env
+					if actualLocalHostAddrIPv6 != "" {
+						if bind == actualLocalHostAddrIPv4 {
+							listenerOpts.extraBind = []string{actualLocalHostAddrIPv6}


do we want to make bind and extraBind? Should we just have bind be a list and we automatically make them "extra"?

In envoy, is there any difference between the primary address and the "extra" addresses?

do we want to make bind and extraBind? Should we just have bind be a list and we automatically make them "extra"?

In envoy, is there any difference between the primary address and the "extra" addresses?

Hi @howardjohn, as discussing in previous PR, ClusterVIPs can holds all IP instances which is available both for bind and extraBind.

Yes, according to the issue envoyproxy/envoy#11184 in envoy, additinal_addresses can hold all other IP instances for one specified listener, except the primary one. And here is the API definition for it.

@howardjohn For example, if there are 2 IP instances ["127.0.0.1", "::1"], then the primary address should be 127.0.0.1, and the extra addresses should be [::1]

I get that's how it's implemented but my question is why? Do we really need to carry around a "address" and "extra addresses" or just a list of addresses and stick a random one as the main on in envoy config?

My concern with this PR, as noted before, is not the generates XDS but the large complexity we are adding to the code for something relatively simple, and lack of integration tests.

If we need all of these complex if statements and juggling around 4 address (localhost + wildcard x v4 v6) it's complex and error prone. Maybe this logic can be factored out into a struct, etc

Hi @howardjohn, how about to factor the extra addresses adding logic into the function of structs buildListenerOpts and inboundChainConfig, then the similar if statements, such as if len(cc.extraBind) > 0 && util.IsIstioVersionGE116(lb.node.IstioVersion) will be eliminated.
For integration test, the dual stack integration test will be added by @jacob-delgado

I get that's how it's implemented but my question is why? Do we really need to carry around a "address" and "extra addresses" or just a list of addresses and stick a random one as the main on in envoy config?
Hi @howardjohn, for this part, current solution has the similar implementation with the API definition of envoy: one primary address and some extra addresses. So it should be more reasonable.

We don't need to make all of our code much more complex just because envoy has a weird API, imo. We should pick the right API for our usage and then translate it to envoy as needed

Integration tests are an expectation for all features in Istio; please have them alongside the PR instead of afterwards. We can wait. If you want to merge without integration tests it can be done in the experimental branch

We don't need to make all of our code much more complex just because envoy has a weird API, imo. We should pick the right API for our usage and then translate it to envoy as needed

Sure, agree with this idea, I will make the code as simple as possible, then re-submit it.

Integration tests are an expectation for all features in Istio; please have them alongside the PR instead of afterwards. We can wait. If you want to merge without integration tests it can be done in the experimental branch

@howardjohn Sure, I will add integration tests for this feature, and make sure the additional_addresses can be correctly added. Thanks for the comment here.

zhlsunshine · 2022-09-05T14:10:32Z

Hi @howardjohn, could you please help to approve this PR if it's okay? Thanks! ^_^

hzxuzhonghu

Do we have integration test which has dual nics?

zhlsunshine · 2022-09-06T04:38:40Z

Do we have integration test which has dual nics?

Hi @hzxuzhonghu , that's a good question, in fact @jacob-delgado is focusing on building the integration test environment for dual stack.

howardjohn · 2022-09-23T14:47:26Z

The kb/msg metric is concerning:

name                                              old kb/msg         new kb/msg         delta
pkg:istio.io/istio/pilot/pkg/xds goos:linux goarch:amd64
RouteGeneration/gateways-8                               1.16k ± 0%         1.16k ± 0%      ~     (all equal)
RouteGeneration/gateways-shared-8                        5.37k ± 0%         5.37k ± 0%      ~     (all equal)
RouteGeneration/empty-8                                   38.9 ± 0%          77.8 ± 0%   +99.77%  (p=0.008 n=5+5)
RouteGeneration/tls-8                                     38.9 ± 0%          77.8 ± 0%   +99.77%  (p=0.008 n=5+5)
RouteGeneration/telemetry-8                               38.9 ± 0%          77.8 ± 0%   +99.77%  (p=0.008 n=5+5)
RouteGeneration/telemetry-api-8                           38.9 ± 0%          77.8 ± 0%   +99.77%  (p=0.008 n=5+5)
RouteGeneration/virtualservice-8                          96.6 ± 0%          96.6 ± 0%      ~     (all equal)
RouteGeneration/authorizationpolicy-8                     0.41 ± 0%          0.41 ± 0%      ~     (all equal)
RouteGeneration/peerauthentication-8                      0.41 ± 0%          0.41 ± 0%      ~     (all equal)
RouteGeneration/knative-gateway-8                          258 ± 0%           258 ± 0%      ~     (all equal)
RouteGeneration/serviceentry-workloadentry-8              38.9 ± 0%          77.8 ± 0%   +99.77%  (p=0.008 n=5+5)
ClusterGeneration/gateways-8                               429 ± 0%           429 ± 0%      ~     (all equal)
ClusterGeneration/gateways-shared-8                        429 ± 0%           429 ± 0%      ~     (all equal)
ClusterGeneration/empty-8                                  375 ± 0%           376 ± 0%    +0.08%  (p=0.008 n=5+5)
ClusterGeneration/tls-8                                    337 ± 0%           338 ± 0%    +0.09%  (p=0.008 n=5+5)
ClusterGeneration/telemetry-8                              188 ± 0%           189 ± 0%    +0.11%  (p=0.008 n=5+5)
ClusterGeneration/telemetry-api-8                          188 ± 0%           189 ± 0%    +0.11%  (p=0.008 n=5+5)
ClusterGeneration/virtualservice-8                        43.4 ± 0%          43.6 ± 0%    +0.58%  (p=0.008 n=5+5)
ClusterGeneration/authorizationpolicy-8                   1.49 ± 0%          1.75 ± 0%   +16.96%  (p=0.008 n=5+5)
ClusterGeneration/peerauthentication-8                    1.49 ± 0%          1.75 ± 0%   +16.96%  (p=0.008 n=5+5)
ClusterGeneration/knative-gateway-8                       1.13 ± 0%          1.13 ± 0%      ~     (all equal)
ClusterGeneration/serviceentry-workloadentry-8             395 ± 0%           395 ± 0%    +0.08%  (p=0.008 n=5+5)
ListenerGeneration/gateways-8                            2.09k ± 0%         2.10k ± 0%    +0.29%  (p=0.008 n=5+5)
ListenerGeneration/gateways-shared-8                      4.00 ± 0%          4.03 ± 0%    +0.75%  (p=0.008 n=5+5)
ListenerGeneration/empty-8                                18.5 ± 0%          46.6 ± 0%  +152.41%  (p=0.008 n=5+5)
ListenerGeneration/tls-8                                  12.6 ± 0%          38.2 ± 0%  +202.54%  (p=0.008 n=5+5)
ListenerGeneration/telemetry-8                            29.1 ± 0%         108.1 ± 0%  +271.48%  (p=0.008 n=5+5)
ListenerGeneration/telemetry-api-8                        54.7 ± 0%         232.4 ± 0%  +325.10%  (p=0.008 n=5+5)
ListenerGeneration/virtualservice-8                       7.32 ± 0%         13.16 ± 0%   +79.90%  (p=0.008 n=5+5)
ListenerGeneration/authorizationpolicy-8                  56.4 ± 0%         109.9 ± 0%   +95.00%  (p=0.008 n=5+5)
ListenerGeneration/peerauthentication-8                   6.41 ± 0%          9.83 ± 0%   +53.45%  (p=0.008 n=5+5)
ListenerGeneration/knative-gateway-8                      3.56 ± 0%          3.59 ± 0%    +0.84%  (p=0.008 n=5+5)
ListenerGeneration/serviceentry-workloadentry-8           18.5 ± 0%          46.6 ± 0%  +152.41%  (p=0.008 n=5+5)

Why are we seeing such a huge increase in config size?

zhlsunshine · 2022-09-26T03:13:24Z

Hi @howardjohn, thanks for your comparison above, then I see now. and I also checked the comparison between1557747988594429952 by ramaraochavali at Aug 11 23:17:51 and 1572987049097564160 by howardjohn at Sep 23 00:37:35.
As you commented here: #40634 (comment), I think you are correct that DiscoverIPMode called in pilot/pkg/xds/bench_test.go causes such huge increase in config size.
Moreover, I also checked the comparison between 1572987049097564160 by howardjohn at Sep 23 00:37:35 and 1573129384842235904 by zhlsunshine at Sep 23 09:58:00, there is no huge increase for them.

According to current analysis, this PR does not cause the increasing in configuration size if it drops the bech_test's DiscoverIPMode calling. However, in fact, it confused me that DiscoverIPMode calling causes such huge increase in configuration size. You know it just sets the IP mode and GlobalUnicastIP of the proxy.

howardjohn · 2022-09-26T14:11:37Z

/test benchmark

howardjohn · 2022-09-26T21:43:28Z

pilot/pkg/networking/core/v1alpha3/listener_address.go

+		wcLh.localHostIPv4 = LocalhostAddress
+		wcLh.localHostIPv6 = LocalhostIPv6Address
+	default:
+		panic("Unknow IP mode")


Suggested change

panic("Unknow IP mode")

panic("Unknown IP mode")

howardjohn · 2022-09-26T21:45:25Z

pilot/pkg/networking/core/v1alpha3/listener.go

 			bind := egressListener.IstioListener.Bind
 			if bind == "" {
 				if bindToPort {
-					bind = actualLocalHostAddress
+					bind = actualLocalHosts[0]


Why do we need extraBind at all? Instead of bind being a list?

I am very worried that somewhere else we will assume there is one bind and mutate it.

For example:

bind := 127.0.0.1 extraBind := ::1 ... lot of code later .. bind = SomeOtherIp // Oops! forgot to change extrabind!

howardjohn · 2022-09-26T21:46:14Z

pilot/pkg/networking/core/v1alpha3/listener.go

 				bindToPort: bindToPort,
 			}
+			if len(extraBind) > 0 {
+				listenerOpts.extraBind = append(listenerOpts.extraBind, extraBind...)
+			}


Suggested change

bindToPort: bindToPort,

}

if len(extraBind) > 0 {

listenerOpts.extraBind = append(listenerOpts.extraBind, extraBind...)

}

bindToPort: bindToPort,

extraBind: extraBind,

}

howardjohn · 2022-09-26T21:46:53Z

pilot/pkg/networking/util/util.go

+		}
+		return
+	}
+	log.Warnf("warning: No extra addresses need to be added to listener [%s]", ll.Name)


Remove this

ramaraochavali

Generally LGTM. Couple of nits. Still not clear what is the issue with bench marks though

ramaraochavali · 2022-09-28T09:05:21Z

pilot/pkg/model/context.go

+	// wildCardIPs contains the wildcard host IPs, including IPv4 and IPv6
+	wildCardIPs []string
+
+	// localHostIPs contains the wildcard host IPs, including IPv4 and IPv6


nit: localHostIPs contains the local host IPs,

ramaraochavali · 2022-09-28T09:08:32Z

pilot/pkg/networking/core/v1alpha3/listener.go

 			bind := egressListener.IstioListener.Bind
 			if bind == "" {
 				if bindToPort {
-					bind = actualLocalHostAddress
+					bind = actualLocalHosts[0]


can you please add a comment on why we are using bind = actualLocalHosts[0]?

hzxuzhonghu

One question, when user specify Bind explicitly, will we add additional Bind now?

hzxuzhonghu · 2022-09-29T03:48:06Z

pilot/pkg/model/context.go

 	}
 	node.GlobalUnicastIP = networkutil.GlobalUnicastIP(node.IPAddresses)
+	node.wildCardIPs = wildCardIPs
+	node.localHostIPs = localHostIPs


I donot feel very comfortable caching same or limited number of strings per proxy. From memory view, this is not friendly

Hi @hzxuzhonghu, in fact they are not caching same, the content for wildCardIPs and localHostIPs depends on the service deployment, it can be, for example to localHostIPs, ["127.0.0.1"], ["::1"] or ["127.0.0.1", "::1"].

BTW,

if not store them in proxy, then Istio must calculate them,

if not calculate them, then must store them
One of them should be choose.

Hi @hzxuzhonghu, use 2 map to maintain wildCardIPs and localHostIPs now

hzxuzhonghu · 2022-09-29T03:48:52Z

pilot/pkg/model/context.go

+// Wildcards returns the wildCard IP addresses for current proxy
+func (node *Proxy) Wildcards() []string {
+	if len(node.wildCardIPs) == 0 {
+		node.DiscoverIPMode()


no call it every, this can be called in init proxy

Hi @hzxuzhonghu, yeah, DiscoverIPMode would be called one time if len(node.wildCardIPs) == 0 or len(node.localHosts) == 0.
In fact, I'm afraid that DiscoverIPMode is not called, so just in case. Do you suggest that remove the calling in both Wildcards() and Localhosts()?

hzxuzhonghu · 2022-09-29T03:52:41Z

pilot/pkg/networking/core/v1alpha3/listener.go

+					bind = actualWildcards[0]
+					if len(actualWildcards) > 1 {
+						extraBind = actualWildcards[1:]
+					}
 				}


Can we cache the result of bind and extraBind or quickly inferred somewhere

Hi @hzxuzhonghu, I think this may be the same as above answer of mine.

hzxuzhonghu · 2022-09-29T03:56:02Z

pilot/pkg/networking/core/v1alpha3/listener.go

@@ -389,12 +401,13 @@ func (lb *ListenerBuilder) buildSidecarOutboundListeners(node *model.Proxy,
 				bind:       bind,
 				port:       listenPort,
 				bindToPort: bindToPort,
+				extraBind:  extraBind,


how about additionalBind

the extraBind will be translated into additional addresses for listener in later.

hzxuzhonghu · 2022-09-29T03:59:39Z

pilot/pkg/networking/util/util.go

@@ -173,6 +173,22 @@ func BuildAddress(bind string, port uint32) *core.Address {
 	}
 }

+// BuildExtraAddresses can add extra addresses to additional addresses for a listener
+func BuildExtraAddresses(extrAddresses []string, listenPort uint32, ll *listener.Listener, node *model.Proxy) {


Suggest moving listener param here, this function should be simple

listener.AdditionalAddresses = BuildExtraAddresses()

okay, I understand, generate the additional addresses in BuildExtraAddresses and remove the listener parameter, then specify as listener.AdditionalAddresses = BuildExtraAddresses()

zhlsunshine · 2022-09-29T05:31:30Z

One question, when user specify Bind explicitly, will we add additional Bind now?

Hi @hzxuzhonghu, what do you mean specify Bind explicitly? the additional bind is added when istio service has more than one addresses.

hzxuzhonghu · 2022-09-29T06:12:50Z

Not sure this is right in this case

zhlsunshine · 2022-09-29T06:25:16Z

Not sure this is right in this case

Hi @hzxuzhonghu, I re-submitted, PTAL, thanks for your comments! ^_^

zhlsunshine · 2022-09-29T08:19:17Z

/test integ-security-multicluster

…de of proxy

zhlsunshine · 2022-09-29T11:11:45Z

Hi @hzxuzhonghu, re-submitted, maintain 2 maps to fetch host addresses according to proxy’s IP mode. Thanks for your comments! ^_^

hzxuzhonghu · 2022-09-30T02:26:47Z

pilot/pkg/networking/util/util.go

@@ -173,6 +173,23 @@ func BuildAddress(bind string, port uint32) *core.Address {
 	}
 }

+// BuildExtraAddresses can add extra addresses to additional addresses for a listener
+func BuildExtraAddresses(extrAddresses []string, listenPort uint32, node *model.Proxy) []*listener.AdditionalAddress {


replace Extra with Additional

okay, no problem!

…alAddresses

add extra addresses for listeners configuration to support dual stack…

d214261

… in Istio

zhlsunshine requested a review from a team as a code owner August 24, 2022 13:58

istio-testing added the size/L Denotes a PR that changes 100-499 lines, ignoring generated files. label Aug 24, 2022

zhlsunshine added the release-notes-none Indicates a PR that does not require release notes. label Aug 24, 2022

howardjohn reviewed Aug 25, 2022

View reviewed changes

zhlsunshine added 3 commits August 26, 2022 16:22

Merge branch 'master' of https://github.com/istio/istio into extraddrs

7ae736c

Merge branch 'master' of https://github.com/istio/istio into extraddrs

4c55384

change code based on comments

9b1a340

zhlsunshine requested a review from a team as a code owner August 29, 2022 07:28

zhlsunshine added 2 commits August 29, 2022 18:19

fixing bug for inboundCustomListener

5bba66c

fixing bug for lds and inbound, and add unit test

291f353

istio-testing added size/XL Denotes a PR that changes 500-999 lines, ignoring generated files. and removed size/L Denotes a PR that changes 100-499 lines, ignoring generated files. labels Aug 30, 2022

refactor some code and fix the unit test error

67c70ed

istio-testing added size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files. and removed size/XL Denotes a PR that changes 500-999 lines, ignoring generated files. labels Aug 30, 2022

zhlsunshine added 2 commits August 30, 2022 23:28

Merge branch 'master' of https://github.com/istio/istio into extraddrs

8cd3827

fix lint-go

57ab29c

howardjohn reviewed Sep 2, 2022

View reviewed changes

zhlsunshine added 2 commits September 5, 2022 11:02

Merge branch 'master' of https://github.com/istio/istio into extraddrs

7bdee39

code change based on comment

d77d11e

istio-testing added size/XL Denotes a PR that changes 500-999 lines, ignoring generated files. and removed size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files. labels Sep 5, 2022

hzxuzhonghu reviewed Sep 6, 2022

View reviewed changes

change code based on comment

20b9b73

Merge branch 'master' of https://github.com/istio/istio into extraddrs

50238c7

howardjohn reviewed Sep 26, 2022

View reviewed changes

zhlsunshine added 3 commits September 27, 2022 08:53

refactor code based on the comments

4143e75

Merge branch 'master' of https://github.com/istio/istio into extraddrs

7648eaa

move the HostAddresses into proxy node according to comments

a6986b1

istio-testing added size/XL Denotes a PR that changes 500-999 lines, ignoring generated files. and removed size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files. labels Sep 28, 2022

ramaraochavali reviewed Sep 28, 2022

View reviewed changes

added comments and fixed nit

04f37e9

hzxuzhonghu reviewed Sep 29, 2022

View reviewed changes

zhlsunshine added 2 commits September 29, 2022 15:22

change the code based on maintainer's comments

e27ee98

Merge branch 'master' of https://github.com/istio/istio into extraddrs

d6c3648

maintain 2 maps to return wildCards and localHosts according to IP mo…

45b5a97

…de of proxy

hzxuzhonghu reviewed Sep 30, 2022

View reviewed changes

hzxuzhonghu approved these changes Sep 30, 2022

View reviewed changes

zhlsunshine added 2 commits September 30, 2022 11:04

change util function name from BuildExtraAddresses into BuildAddition…

0cdd8aa

…alAddresses

Merge branch 'master' of https://github.com/istio/istio into extraddrs

9293b01

istio-testing merged commit 0812983 into istio:master Sep 30, 2022

zhlsunshine mentioned this pull request Dec 15, 2022

add zhlsunshine as Networking Maintainers/Pilot maintainer istio/community#915

Closed

5 tasks

zhlsunshine mentioned this pull request Apr 24, 2024

Add IPv4/IPv6 Dual Stack Support envoyproxy/gateway#184

Closed

	bind = egressListener.IstioListener.Bind
	bind := egressListener.IstioListener.Bind

Add extra addresses for listeners configuration to support dual stack in Istio #40634

Add extra addresses for listeners configuration to support dual stack in Istio #40634

Uh oh!

Conversation

zhlsunshine commented Aug 24, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jacob-delgado commented Aug 24, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

howardjohn left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

hzxuzhonghu commented Aug 30, 2022

Uh oh!

zhlsunshine commented Aug 30, 2022

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zhlsunshine Sep 5, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zhlsunshine commented Sep 5, 2022

Uh oh!

hzxuzhonghu left a comment

Choose a reason for hiding this comment

Uh oh!

zhlsunshine commented Sep 6, 2022

Uh oh!

howardjohn commented Sep 23, 2022

Uh oh!

zhlsunshine commented Sep 26, 2022

Uh oh!

howardjohn commented Sep 26, 2022

Uh oh!

zhlsunshine commented Aug 24, 2022 •

edited

Loading

jacob-delgado commented Aug 24, 2022 •

edited

Loading

zhlsunshine Sep 5, 2022 •

edited

Loading

zhlsunshine Sep 29, 2022 •

edited

Loading