Add support for injecting jaeger agent in annotated statefulsets. #78

annanay25 · 2018-11-01T19:21:48Z

Attempts to resolve #62.

Need to check test coverage.

jpkrohling · 2018-11-01T19:21:54Z

This change is

codecov · 2018-11-02T07:20:44Z

Codecov Report

Merging #78 into master will increase coverage by 7.89%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #78      +/-   ##
==========================================
+ Coverage   88.55%   96.45%   +7.89%     
==========================================
  Files          70       30      -40     
  Lines        3119     1466    -1653     
==========================================
- Hits         2762     1414    -1348     
+ Misses        244       40     -204     
+ Partials      113       12     -101

Impacted Files	Coverage Δ
pkg/inject/sidecar.go	`100% <100%> (ø)`	⬆️
pkg/cronjob/es_index_cleaner.go	`96% <0%> (-4%)`	⬇️
pkg/config/ui/ui.go	`96.42% <0%> (-0.45%)`	⬇️
pkg/config/sampling/sampling.go	`96.49% <0%> (-0.44%)`	⬇️
pkg/cronjob/spark_dependencies.go	`97.46% <0%> (-0.31%)`	⬇️
pkg/storage/dependency.go	`100% <0%> (ø)`	⬆️
pkg/strategy/controller.go	`100% <0%> (ø)`	⬆️
pkg/account/oauth-proxy.go	`100% <0%> (ø)`	⬆️
pkg/route/query.go	`100% <0%> (ø)`	⬆️
pkg/deployment/collector.go	`100% <0%> (ø)`	⬆️
... and 56 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 44208d9...5a694fd. Read the comment docs.

jpkrohling · 2018-11-02T11:29:35Z

Let me know once it's ready to be reviewed.

annanay25 · 2018-11-09T06:55:40Z

@jpkrohling request for review.

jpkrohling

Reviewed 5 of 5 files at r1.
Reviewable status: all files reviewed, 4 unresolved discussions (waiting on @annanay25)

a discussion (no related file):
I like the idea of getting the automatic sidecar injection to work for stateful sets, and I'm thankful for your efforts!

I think the approach from this PR won't scale when we think about other Pod container types. A better plan would be to make the current code more generic, to accept sdk.Object instead of Deployment. This way, a good part of the code could be reused for all types. Custom behavior can be achieved through private, specialized functions.

I have a couple of other things that should be considered before merging:

Documentation -- the README would be outdated with this PR
e2e test -- we already have one for the current feature, under test/e2e/sidecar.go. You could rename it to sidecar-deployment.go and create another one for sidecar-statefulset.go. Don't worry about duplication of code here in the tests: in most cases, it's actually desirable to have duplicated code here, so that the tests are very explicit about what they assume

pkg/inject/sidecar.go, line 74 at r1 (raw file):

// Return a container for sidecar injection
func Container(jaeger *v1alpha1.Jaeger) v1.Container {

I would prefer to keep this private, leaving the Sidecar function as the single entry point to this kind of feature. You could move the existing Inject function to, say, injectIntoDeployment and create a new function named Inject(sdk.Object, *v1alpha1.Jaeger). The sdk.Object is then type-checked and the appropriate specialized injectXXX is called. Your new feature would then be inside a injectIntoStatefulSet.

pkg/statefulset/agent.go, line 1 at r1 (raw file):

package statefulset

This is mostly a duplicate of what we have under the inject package, sidecar.go source file. It would probably not be hard to change that to accommodate StatefulSets.

pkg/stub/handler.go, line 92 at r1 (raw file):

			}
		}
	case *appsv1.StatefulSet:

This also seems like a good duplication of the logic for the Deployment part. It would probably be possible to make the same code work for both cases, especially if the Needed and Select functions were changed to accept a sdk.Object instead of Deployment.

annanay25 · 2018-11-10T20:14:29Z

Thanks for the comprehensive review!

I understand that we need to write sidecar injection for multiple pod types, I am looking into making the inject package more generic by accepting all deployment types. Will update this PR soon.

annanay25

Package statefulset has been deleted, and its functionality written within package inject.
Functions Sidecar, Needed, Select have been modified to handle sdk.Object.

Todo:

Add tests
Update README

pkg/stub/handler.go

jpkrohling · 2018-11-12T13:25:11Z

@annanay25 , let me know once this is ready for a new review round. Currently, the unit tests are failing, so, I assume it's not ready yet.

annanay25 · 2018-11-12T19:10:35Z

@jpkrohling I was a little confused as to how to proceed here - #78 (comment)

Your thoughts?

jpkrohling · 2018-11-13T15:38:48Z

Your thoughts?

Sorry, I missed that part. I just replied with a suggestion that might work.

jpkrohling · 2018-11-14T08:33:42Z

Heads up: I'm starting to work on #66, which will probably conflict with this PR. Please, wait a couple of days to work on this issue, as I'm not sure how much of it would be reusable there.

annanay25 · 2018-11-14T18:44:16Z

Understood. I'll hold this off, until later. Thanks!

jpkrohling · 2018-11-20T15:13:10Z

The migration to the new operator SDK is now complete. Please, rebase and update this PR.

annanay25

Updated the PR.

Todo: Tests for StatefulSet

pkg/inject/sidecar.go

pkg/controller/deployment/deployment_controller.go

annanay25 · 2018-12-06T20:02:40Z

@jpkrohling the case statement grouping as implemented here - https://github.com/jaegertracing/jaeger-operator/pull/78/files#diff-f88c99835d4b2acda532e36123761825R91 - is not evaluating the *appsv1.Deployment section (its breaking out without running the part under *appsv1.StatefulSet)

annanay25 · 2018-12-11T18:31:38Z

@jpkrohling I updated this patch, but there's a bunch of replicated code in the inject package which I'm not sure how to workaround.

jpkrohling

Reviewed 1 of 4 files at r2, 1 of 5 files at r3, 3 of 4 files at r4.
Reviewable status: all files reviewed, 11 unresolved discussions (waiting on @annanay25 and @jpkrohling)

pkg/inject/sidecar.go, line 44 at r4 (raw file):

			logrus.Debugf("Skipping sidecar injection for instance %v", o.Name)
		} else {
			decorate(o)

Can't you collapse this into the case above? Like you did for the deployment_controller.go?

pkg/inject/sidecar.go, line 52 at r4 (raw file):

// Needed determines whether a pod needs to get a sidecar injected or not
func Needed(Name string, Annotations map[string]string, Containers []v1.Container) bool {

I'm curious as to why the var names are capitalized here. Also, it would probably be better to receive an obj like the other functions, do a type check and log in case the type isn't a supported type (Deployment/Statefulset)

pkg/inject/sidecar.go, line 70 at r4 (raw file):

// Select a suitable Jaeger from the JaegerList for the given Pod, or nil of none is suitable
func Select(Annotations map[string]string, availableJaegerPods *v1alpha1.JaegerList) *v1alpha1.Jaeger {

Same question as above

pkg/inject/sidecar.go, line 120 at r4 (raw file):

func decorate(obj runtime.Object) {
	switch o := obj.(type) {
	case *appsv1.Deployment:

Is the code for Deployment exactly like the one for StatefulSet? If not, please leave a comment explaining the difference. If they are the same, then collapse the two cases as you did before in the deployment_controller.go.

annanay25

Reviewable status: all files reviewed, 11 unresolved discussions (waiting on @annanay25 and @jpkrohling)

pkg/inject/sidecar.go, line 44 at r4 (raw file):

Previously, jpkrohling (Juraci Paixão Kröhling) wrote…

Can't you collapse this into the case above? Like you did for the deployment_controller.go?

That doesn't seem to work, and apparently we need to perform complete type checking at least once.

switch instance := obj.(type) {
    case *appsv1.Deployment:
    case *appsv1.StatefulSet:
        decorate(instance)

^ seems to fail for me.

pkg/inject/sidecar.go, line 52 at r4 (raw file):

Previously, jpkrohling (Juraci Paixão Kröhling) wrote…

I'm curious as to why the var names are capitalized here. Also, it would probably be better to receive an obj like the other functions, do a type check and log in case the type isn't a supported type (Deployment/Statefulset)

Capitalisation was erroneous, will change this. You had suggested earlier that it would be better to pass only the necessary variables to functions like Needed and Select - #78 (comment)

pkg/inject/sidecar.go, line 120 at r4 (raw file):

Previously, jpkrohling (Juraci Paixão Kröhling) wrote…

Is the code for Deployment exactly like the one for StatefulSet? If not, please leave a comment explaining the difference. If they are the same, then collapse the two cases as you did before in the deployment_controller.go.

They're the same, but some type checking constraints are forcing me to publish the functions this way.

jpkrohling

Reviewable status: all files reviewed, 11 unresolved discussions (waiting on @annanay25 and @jpkrohling)

pkg/inject/sidecar.go, line 44 at r4 (raw file):

Previously, annanay25 (Annanay Agarwal) wrote…

That doesn't seem to work, and apparently we need to perform complete type checking at least once.
switch instance := obj.(type) {
    case *appsv1.Deployment:
    case *appsv1.StatefulSet:
        decorate(instance) 
^ seems to fail for me.

This seems to work:

func Sidecar(obj runtime.Object, jaeger *v1alpha1.Jaeger) {
	deployment.NewAgent(jaeger) // we need some initialization from that, but we don't actually need the agent's instance here

	switch o := obj.(type) {
	case *appsv1.Deployment:
	case *appsv1.StatefulSet:
		if jaeger == nil || o.Annotations[Annotation] != jaeger.Name {
			logrus.Debugf("Skipping sidecar injection for instance %v", o.Name)
		} else {
			decorate(o)
			logrus.Debugf("Injecting sidecar for pod %v", o.Name)
			o.Spec.Template.Spec.Containers = append(o.Spec.Template.Spec.Containers, container(jaeger))
		}
	}
}

pkg/inject/sidecar.go, line 52 at r4 (raw file):

Previously, annanay25 (Annanay Agarwal) wrote…

Capitalisation was erroneous, will change this. You had suggested earlier that it would be better to pass only the necessary variables to functions like Needed and Select - #78 (comment)

I'd have to go back to see why I suggested that. Was it perhaps because you'd need to pass a Go generic object (interface{}), instead of Kubernetes'? It's certainly better to pass name/annotations/containers instead of a generic Go object, but I'd rather make it accept a Kubernetes runtime object if possible.

pkg/inject/sidecar.go, line 120 at r4 (raw file):

Previously, annanay25 (Annanay Agarwal) wrote…

They're the same, but some type checking constraints are forcing me to publish the functions this way.

Looks like this might work:

func decorate(obj runtime.Object) {
	switch o := obj.(type) {
	case *appsv1.Deployment:
	case *appsv1.StatefulSet:
		if app, found := o.Spec.Template.Labels["app"]; found {
			// Append the namespace to the app name. Using the DNS style "<app>.<namespace>""
			// which also matches with the style used in Istio.
			if len(o.Namespace) > 0 {
				app += "." + o.Namespace
			} else {
				app += ".default"
			}
			for i := 0; i < len(o.Spec.Template.Spec.Containers); i++ {
				if !hasEnv(envVarServiceName, o.Spec.Template.Spec.Containers[i].Env) {
					o.Spec.Template.Spec.Containers[i].Env = append(o.Spec.Template.Spec.Containers[i].Env, v1.EnvVar{
						Name:  envVarServiceName,
						Value: app,
					})
				}
				if !hasEnv(envVarPropagation, o.Spec.Template.Spec.Containers[i].Env) {
					o.Spec.Template.Spec.Containers[i].Env = append(o.Spec.Template.Spec.Containers[i].Env, v1.EnvVar{
						Name:  envVarPropagation,
						Value: "jaeger,b3",
					})
				}
			}
		}
	}
}

annanay25

Reviewable status: all files reviewed, 11 unresolved discussions (waiting on @annanay25 and @jpkrohling)

pkg/inject/sidecar.go, line 44 at r4 (raw file):

Previously, jpkrohling (Juraci Paixão Kröhling) wrote…

This seems to work:

func Sidecar(obj runtime.Object, jaeger *v1alpha1.Jaeger) {
	deployment.NewAgent(jaeger) // we need some initialization from that, but we don't actually need the agent's instance here

	switch o := obj.(type) {
	case *appsv1.Deployment:
	case *appsv1.StatefulSet:
		if jaeger == nil || o.Annotations[Annotation] != jaeger.Name {
			logrus.Debugf("Skipping sidecar injection for instance %v", o.Name)
		} else {
			decorate(o)
			logrus.Debugf("Injecting sidecar for pod %v", o.Name)
			o.Spec.Template.Spec.Containers = append(o.Spec.Template.Spec.Containers, container(jaeger))
		}
	}
}

When I make this change, the following happens:

➜  jaeger-operator git:(statefulsets-sidecar) ✗ make test
Running unit tests...
?   	github.com/jaegertracing/jaeger-operator/cmd	[no test files]
?   	github.com/jaegertracing/jaeger-operator/cmd/manager	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/account	0.289s	coverage: 100.0% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/apis	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/apis/io/v1alpha1	0.396s	coverage: 17.0% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/cmd/start	[no test files]
?   	github.com/jaegertracing/jaeger-operator/pkg/cmd/version	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/config/sampling	0.278s	coverage: 94.1% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/config/ui	0.356s	coverage: 94.1% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/controller	[no test files]
?   	github.com/jaegertracing/jaeger-operator/pkg/controller/deployment	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/controller/jaeger	3.567s	coverage: 64.3% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/cronjob	0.341s	coverage: 93.5% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/deployment	0.288s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/ingress	0.205s	coverage: 100.0% of statements
--- FAIL: TestInjectSidecar (0.00s)
	sidecar_test.go:33:
			Error Trace:	sidecar_test.go:33
			Error:      	"[{  [] []  [] [] [] {map[] map[]} [] [] nil nil nil    nil %!s(bool=false) %!s(bool=false) %!s(bool=false)}]" should have 2 item(s), but has 1
			Test:       	TestInjectSidecar
panic: runtime error: index out of range [recovered]
	panic: runtime error: index out of range

goroutine 11 [running]:
testing.tRunner.func1(0xc42016a960)
	/usr/local/go/src/testing/testing.go:742 +0x29d
panic(0x1753be0, 0x1c9e3f0)
	/usr/local/go/src/runtime/panic.go:502 +0x229
github.com/jaegertracing/jaeger-operator/pkg/inject.TestInjectSidecar(0xc42016a960)
	/Users/annanay.a/Desktop/git/go/src/github.com/jaegertracing/jaeger-operator/pkg/inject/sidecar_test.go:34 +0x2e0
testing.tRunner(0xc42016a960, 0x1865c88)
	/usr/local/go/src/testing/testing.go:777 +0xd0
created by testing.(*T).Run
	/usr/local/go/src/testing/testing.go:824 +0x2e0
FAIL	github.com/jaegertracing/jaeger-operator/pkg/inject	0.293s
ok  	github.com/jaegertracing/jaeger-operator/pkg/route	0.405s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/service	0.275s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/storage	0.252s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/strategy	0.265s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/util	0.212s	coverage: 100.0% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/version	[no test files]
make: *** [unit-tests] Error 1

... because it finds the body of case *appsv1.Deployment: empty and breaks out without doing anything.

jpkrohling

Reviewable status: all files reviewed, 11 unresolved discussions (waiting on @annanay25 and @jpkrohling)

pkg/inject/sidecar.go, line 44 at r4 (raw file):

Previously, annanay25 (Annanay Agarwal) wrote…

When I make this change, the following happens:

➜  jaeger-operator git:(statefulsets-sidecar) ✗ make test
Running unit tests...
?   	github.com/jaegertracing/jaeger-operator/cmd	[no test files]
?   	github.com/jaegertracing/jaeger-operator/cmd/manager	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/account	0.289s	coverage: 100.0% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/apis	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/apis/io/v1alpha1	0.396s	coverage: 17.0% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/cmd/start	[no test files]
?   	github.com/jaegertracing/jaeger-operator/pkg/cmd/version	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/config/sampling	0.278s	coverage: 94.1% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/config/ui	0.356s	coverage: 94.1% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/controller	[no test files]
?   	github.com/jaegertracing/jaeger-operator/pkg/controller/deployment	[no test files]
ok  	github.com/jaegertracing/jaeger-operator/pkg/controller/jaeger	3.567s	coverage: 64.3% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/cronjob	0.341s	coverage: 93.5% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/deployment	0.288s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/ingress	0.205s	coverage: 100.0% of statements
--- FAIL: TestInjectSidecar (0.00s)
	sidecar_test.go:33:
			Error Trace:	sidecar_test.go:33
			Error:      	"[{  [] []  [] [] [] {map[] map[]} [] [] nil nil nil    nil %!s(bool=false) %!s(bool=false) %!s(bool=false)}]" should have 2 item(s), but has 1
			Test:       	TestInjectSidecar
panic: runtime error: index out of range [recovered]
	panic: runtime error: index out of range

goroutine 11 [running]:
testing.tRunner.func1(0xc42016a960)
	/usr/local/go/src/testing/testing.go:742 +0x29d
panic(0x1753be0, 0x1c9e3f0)
	/usr/local/go/src/runtime/panic.go:502 +0x229
github.com/jaegertracing/jaeger-operator/pkg/inject.TestInjectSidecar(0xc42016a960)
	/Users/annanay.a/Desktop/git/go/src/github.com/jaegertracing/jaeger-operator/pkg/inject/sidecar_test.go:34 +0x2e0
testing.tRunner(0xc42016a960, 0x1865c88)
	/usr/local/go/src/testing/testing.go:777 +0xd0
created by testing.(*T).Run
	/usr/local/go/src/testing/testing.go:824 +0x2e0
FAIL	github.com/jaegertracing/jaeger-operator/pkg/inject	0.293s
ok  	github.com/jaegertracing/jaeger-operator/pkg/route	0.405s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/service	0.275s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/storage	0.252s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/strategy	0.265s	coverage: 100.0% of statements
ok  	github.com/jaegertracing/jaeger-operator/pkg/util	0.212s	coverage: 100.0% of statements
?   	github.com/jaegertracing/jaeger-operator/pkg/version	[no test files]
make: *** [unit-tests] Error 1

... because it finds the body of case *appsv1.Deployment: empty and breaks out without doing anything.

Sorry, my bad. I should have added a comma and then would realize that it won't work anyway:

https://golang.org/ref/spec#Type_switches

My main problem with the current code is that it won't scale nicely for all types that include a PodSpec/PodTemplateSpec, like batches, jobs, daemon sets, ... How about creating a callback-type of function, that accepts a list of containers? Or perhaps you can come up with a way to have the Sidecar/decorate to accept a PodTemplateSpec? That's pretty much all you need to inject a sidecar, no?

annanay25 · 2019-01-29T19:40:01Z

My main problem with the current code is that it won't scale nicely for all types that include a PodSpec/PodTemplateSpec, like batches, jobs, daemon sets, ..

Completely agree. We need a better way to do this.

I think we need more than the PodTemplateSpec in the decorate function for example, where we're working with the object namespace:

// Append the namespace to the app name. Using the DNS style "<app>.<namespace>""
// which also matches with the style used in Istio.
if len(o.Namespace) > 0 {
    app += "." + o.Namespace
} else {
    app += ".default"
}

jpkrohling · 2019-02-04T12:58:33Z

I think we need more than the PodTemplateSpec in the decorate function for example, where we're working with the object namespace:

That's fine, you can just pass the namespace then. Is that all you'd need? PodTemplateSpec + Namespace?

jpkrohling · 2019-02-28T08:40:23Z

I think there are quite a few merge conflicts now. Would you like to fix them, or should I close it?

annanay25 · 2019-03-04T19:29:08Z

I'm going to take a stab at this soon :)

Signed-off-by: Annanay <[email protected]>

annanay25 · 2019-03-11T19:56:14Z

@jpkrohling could you tell me if the functions look good? I'll correct the tests accordingly :)

jpkrohling · 2019-03-21T13:19:18Z

Looks like the build is failing:

pkg/controller/deployment/deployment_controller.go:90:9: expected switch expression, found assignment (missing parentheses around composite literal?)

jpkrohling

Reviewed 1 of 4 files at r5.
Reviewable status: 1 of 3 files reviewed, 12 unresolved discussions (waiting on @annanay25 and @jpkrohling)

pkg/controller/deployment/deployment_controller.go, line 91 at r5 (raw file):

	switch o := *obj.(type) {
	case *appsv1.Deployment:

Sorry for bringing this up again, but could you try to use v1.Object from https://github.com/kubernetes/apimachinery/blob/master/pkg/apis/meta/v1/meta.go#L33 ?, where v1 is "k8s.io/apimachinery/pkg/apis/meta/v1"

I recently found this one and it might solve this type of problem.

Signed-off-by: Annanay <[email protected]>

annanay25 · 2019-03-23T17:55:30Z

runtime.Object cannot be type casted into meta/v1.Object - I'm missing the point here

jpkrohling · 2019-03-25T12:38:04Z

runtime.Object cannot be type casted into meta/v1.Object - I'm missing the point here

I mean, instead of having almost the same code for appsv1.Deployment and appsv1.StatefulSet, we could try having meta/v1.Object instead. IIRC, we needed something specific for Deployment or Statefulset in only one line, right?

jpkrohling · 2019-07-02T12:56:45Z

@annanay25, I'm closing this one, as we are currently working on the webhook approach (see #171), which would make this obsolete. If you still want to work on this before we move on with that other approach, feel free to reopen.

annanay25 force-pushed the statefulsets-sidecar branch 2 times, most recently from fcb3a91 to 9f64d5f Compare November 2, 2018 07:18

annanay25 changed the title ~~WIP: Add support for injecting jaeger agent in annotated statefulsets.~~ Add support for injecting jaeger agent in annotated statefulsets. Nov 9, 2018

jpkrohling reviewed Nov 9, 2018

View reviewed changes

annanay25 commented Nov 11, 2018

View reviewed changes

pkg/stub/handler.go Outdated Show resolved Hide resolved

annanay25 force-pushed the statefulsets-sidecar branch from 8bb98a0 to 0e28d44 Compare December 4, 2018 11:32

annanay25 commented Dec 5, 2018

View reviewed changes

pkg/inject/sidecar.go Outdated Show resolved Hide resolved

pkg/controller/deployment/deployment_controller.go Outdated Show resolved Hide resolved

annanay25 force-pushed the statefulsets-sidecar branch from ebd9569 to e3a81ee Compare December 11, 2018 18:23

jpkrohling reviewed Jan 15, 2019

View reviewed changes

annanay25 commented Jan 15, 2019

View reviewed changes

jpkrohling reviewed Jan 16, 2019

View reviewed changes

annanay25 commented Jan 21, 2019

View reviewed changes

jpkrohling reviewed Jan 22, 2019

View reviewed changes

Annanay added 3 commits March 12, 2019 00:27

WIP: Rebased, updated.

ab858b7

Signed-off-by: Annanay <[email protected]>

Addressed comments

9f04f41

Signed-off-by: Annanay <[email protected]>

Fix switch case and add tests

9a1e78f

Signed-off-by: Annanay <[email protected]>

annanay25 force-pushed the statefulsets-sidecar branch from e3a81ee to 9a1e78f Compare March 11, 2019 19:02

Annanay added 2 commits March 12, 2019 00:37

Correct package name to v1

3da2039

Signed-off-by: Annanay <[email protected]>

Make changes to Reconcile, Decorate, Sidecar functions

611592f

Signed-off-by: Annanay <[email protected]>

jpkrohling reviewed Mar 21, 2019

View reviewed changes

Annanay added 3 commits March 23, 2019 12:23

Address comments, fix function definitions

91abe71

Signed-off-by: Annanay <[email protected]>

Fix object types and function calls

fee9079

Signed-off-by: Annanay <[email protected]>

Check build status without meta/v1

5a694fd

Signed-off-by: Annanay <[email protected]>

jpkrohling closed this Jul 2, 2019

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add support for injecting jaeger agent in annotated statefulsets. #78

Add support for injecting jaeger agent in annotated statefulsets. #78

annanay25 commented Nov 1, 2018

jpkrohling commented Nov 1, 2018

codecov bot commented Nov 2, 2018 •

edited

Loading

jpkrohling commented Nov 2, 2018

annanay25 commented Nov 9, 2018

jpkrohling left a comment

annanay25 commented Nov 10, 2018

annanay25 left a comment

jpkrohling commented Nov 12, 2018

annanay25 commented Nov 12, 2018

jpkrohling commented Nov 13, 2018

jpkrohling commented Nov 14, 2018

annanay25 commented Nov 14, 2018 •

edited

Loading

jpkrohling commented Nov 20, 2018

annanay25 left a comment •

edited

Loading

annanay25 commented Dec 6, 2018

annanay25 commented Dec 11, 2018

jpkrohling left a comment

annanay25 left a comment

jpkrohling left a comment

annanay25 left a comment

jpkrohling left a comment

annanay25 commented Jan 29, 2019

jpkrohling commented Feb 4, 2019

jpkrohling commented Feb 28, 2019

annanay25 commented Mar 4, 2019

annanay25 commented Mar 11, 2019

jpkrohling commented Mar 21, 2019

jpkrohling left a comment

annanay25 commented Mar 23, 2019

jpkrohling commented Mar 25, 2019

jpkrohling commented Jul 2, 2019

Add support for injecting jaeger agent in annotated statefulsets. #78

Add support for injecting jaeger agent in annotated statefulsets. #78

Conversation

annanay25 commented Nov 1, 2018

jpkrohling commented Nov 1, 2018

codecov bot commented Nov 2, 2018 • edited Loading

Codecov Report

jpkrohling commented Nov 2, 2018

annanay25 commented Nov 9, 2018

jpkrohling left a comment

Choose a reason for hiding this comment

annanay25 commented Nov 10, 2018

annanay25 left a comment

Choose a reason for hiding this comment

jpkrohling commented Nov 12, 2018

annanay25 commented Nov 12, 2018

jpkrohling commented Nov 13, 2018

jpkrohling commented Nov 14, 2018

annanay25 commented Nov 14, 2018 • edited Loading

jpkrohling commented Nov 20, 2018

annanay25 left a comment • edited Loading

Choose a reason for hiding this comment

annanay25 commented Dec 6, 2018

annanay25 commented Dec 11, 2018

jpkrohling left a comment

Choose a reason for hiding this comment

annanay25 left a comment

Choose a reason for hiding this comment

jpkrohling left a comment

Choose a reason for hiding this comment

annanay25 left a comment

Choose a reason for hiding this comment

jpkrohling left a comment

Choose a reason for hiding this comment

annanay25 commented Jan 29, 2019

jpkrohling commented Feb 4, 2019

jpkrohling commented Feb 28, 2019

annanay25 commented Mar 4, 2019

annanay25 commented Mar 11, 2019

jpkrohling commented Mar 21, 2019

jpkrohling left a comment

Choose a reason for hiding this comment

annanay25 commented Mar 23, 2019

jpkrohling commented Mar 25, 2019

jpkrohling commented Jul 2, 2019

codecov bot commented Nov 2, 2018 •

edited

Loading

annanay25 commented Nov 14, 2018 •

edited

Loading

annanay25 left a comment •

edited

Loading