Add s390x support on multiarch docker images #2948

kun-lu20 · 2021-04-21T16:45:24Z

This change is to add s390x support on multiarch Jaeger docker images

Signed-off-by: Kun-Lu kun.lu@ibm.com

Which problem is this PR solving?

Fixes Multiarch docker images #2292

Short description of the changes

modifies CI workflows (ci-all-in-one-build.yml, ci-docker-build.yml, ci-release.yml) to build and publish multi-arch docker images on docker.io and quay.io.
adds s390x support on multi-arch docker images of baseimg, all-in-one and all the Jaeger components, except debug images and Jaeger-Cassandra-Schema whose base image doesn’t have s390x support yet.
uses local registry service to hold:
1. multi-arch baseimg,
2. multi-arch all-in-one image for local integration test.
updates Makefile and scripts/build-all-in-one-image.sh, adds scripts/build-upload-docker-images.sh to use docker buildx to build and publish multi-arch docker images.
removes scripts/upload-all-docker-images.sh since its uploading function has been implemented in scripts/build-upload-docker-images.sh.
removes default value of ARG TARGETARCH in Dockerfiles.
all the building, integration testing, tag computation and publishing procedures are the same as existing ones (including all the multi-arch docker images and docker images which currently don’t support s390x).

codecov · 2021-04-21T16:52:06Z

Codecov Report

Merging #2948 (a62c82f) into master (127293d) will decrease coverage by 0.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #2948      +/-   ##
==========================================
- Coverage   95.94%   95.92%   -0.02%     
==========================================
  Files         236      236              
  Lines       10238    10238              
==========================================
- Hits         9823     9821       -2     
- Misses        345      347       +2     
  Partials       70       70

Impacted Files	Coverage Δ
pkg/config/tlscfg/cert_watcher.go	`92.20% <0.00%> (-2.60%)`	⬇️
cmd/query/app/server.go	`95.83% <0.00%> (-1.39%)`	⬇️
cmd/query/app/static_handler.go	`98.38% <0.00%> (+1.61%)`	⬆️

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 127293d...a62c82f. Read the comment docs.

This change is to add s390x support on multiarch Jaeger docker images Signed-off-by: Kun Lu <kun.lu@ibm.com> Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Signed-off-by: Kun-Lu kun.lu@ibm.com

Ashmita152

Thank you for working on it.

Ashmita152 · 2021-04-22T18:07:52Z

scripts/build-all-in-one-image.sh

+  ## the other possible value, currently, is 'master' (or another branch name)
+  if [[ $BRANCH == v* ]]; then
+      COMPONENT_VERSION=$(echo ${BRANCH} | grep -Po "([\d\.]+)")
+      MAJOR_MINOR=$(echo ${COMPONENT_VERSION} | awk -F. '{print $1"."$2}')


Every jaeger release has major, minor and patch components. This looks like patch isn't taken into consideration.

The COMPONENT_VERSION variable in L71 is actually for patch components, like 1.22.0. I've changed the name of this variable to MAJOR_MINOR_PATCH to avoid ambiguity.

Ashmita152 · 2021-04-22T18:08:38Z

scripts/build-all-in-one-image.sh

+  if [[ $BRANCH == v* ]]; then
+      COMPONENT_VERSION=$(echo ${BRANCH} | grep -Po "([\d\.]+)")
+      MAJOR_MINOR=$(echo ${COMPONENT_VERSION} | awk -F. '{print $1"."$2}')
+      MAJOR1=$(echo ${COMPONENT_VERSION} | awk -F. '{print $1}')


Should we use MAJOR instead of MAJOR1 ?

Yes, I've updated the code here.

Ashmita152 · 2021-04-22T18:12:32Z

scripts/build-all-in-one-image.sh

+  IMAGE_TAGS="${IMAGE_TAGS} --tag docker.io/${SNAPSHOT_TAG}"
+
+  ## for quay.io
+  IMAGE_TAGS="${IMAGE_TAGS} --tag quay.io/${BASE_BUILD_IMAGE} --tag quay.io/${BUILD_IMAGE}"


Do we need to add quay.io tags separately. Can we add as part of L83.

Yes, we can. I've updated the script to combine them.

Ashmita152 · 2021-04-22T18:17:56Z

Makefile

@@ -72,6 +72,17 @@ MOCKERY=mockery

 .DEFAULT_GOAL := test-and-lint

+BASE_IMAGE_MULTIARCH := localhost:5000/baseimg:$(VERSION)-$(shell echo $(ROOT_IMAGE) | tr : -)
+PLATFORMS=linux/amd64,linux/s390x
+INGESTER_TAG=$(subst JAGERCOMP,ingester,$(IMAGE_TAGS))


Do we need to do this for each jaeger component, isn't it the same value after all ? Also where is JAGERCOMP variable populated ?

Yes, I think we need to, since each jaeger component has its unique name, hence the unique tag value.

JAGERCOMP is not a variable but a const string. We use it in the tag computation procedure in scripts/build-upload-docker-images.sh to get the value of variable IMAGE_TAGS. Then in Makefile L77-L84, we replace it with the name of each jaeger component in variable IMAGE_TAGS to get each actual tag value.

I am not a fan of these cross-dependences. If the string that contains JAGERCOMP is defined in the script, why can't the script resolve it as needed, without pushing the substitution into Makefile?

Also, Jaeger is spelled with 'e'.

Got it. I've corrected the spelling and moved the related code to the script.

Ashmita152 · 2021-04-22T18:22:41Z

cmd/all-in-one/Dockerfile

@@ -28,6 +28,8 @@ EXPOSE 16686
 COPY all-in-one-linux-$TARGETARCH /go/bin/all-in-one-linux
 COPY sampling_strategies.json /etc/jaeger/

+RUN chmod +x /go/bin/all-in-one-linux


Why are we adding this chmod here. Isn't the binary already executable ?

Yes, you are right, the binary is already executable. I've removed this line from Dockerfile.

.github/workflows/ci-all-in-one-build.yml

Remove chmod from all-in-one Dockerfile Optimize tag computation code Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-04-23T22:11:57Z

@Ashmita152 Thank you very much for your review comments! I've responded to them and made corresponding code changes. Please take a look when you are available.

yurishkuro · 2021-04-25T22:38:23Z

scripts/build-all-in-one-image.sh

@@ -45,17 +45,96 @@ upload_to_docker() {
  fi
 }

-make build-all-in-one GOOS=linux GOARCH=$GOARCH
+upload_multiarch_to_docker() {


Why is there so much code just for all-in-one? I assume it would be the same for other Docker images, so let's keep a single copy.

Yes, you're right. I've extracted the duplicate code from scripts/build-all-in-one-image.sh and scripts/build-upload-docker-images.sh and put them in newly created scripts.

yurishkuro · 2021-04-25T22:38:34Z

.github/workflows/ci-all-in-one-build.yml

@@ -6,9 +6,14 @@ on:
  pull_request:
    branches: [ master ]

-jobs:
+jobs:   


remove whitespace

Whitespace removed.

.github/workflows/ci-docker-build.yml

yurishkuro · 2021-04-25T22:47:12Z

Makefile

@@ -72,6 +72,17 @@ MOCKERY=mockery

 .DEFAULT_GOAL := test-and-lint

+BASE_IMAGE_MULTIARCH := localhost:5000/baseimg:$(VERSION)-$(shell echo $(ROOT_IMAGE) | tr : -)
+PLATFORMS=linux/amd64,linux/s390x
+INGESTER_TAG=$(subst JAGERCOMP,ingester,$(IMAGE_TAGS))


I am not a fan of these cross-dependences. If the string that contains JAGERCOMP is defined in the script, why can't the script resolve it as needed, without pushing the substitution into Makefile?

Also, Jaeger is spelled with 'e'.

yurishkuro · 2021-04-25T22:48:43Z

Makefile

@@ -348,6 +361,73 @@ docker-images-only: docker-images-cassandra \
 	docker-images-tracegen \
 	docker-images-anonymizer

+.PHONY: multiarch-docker


Couldn't we just move all this code into the script? Why do they need to be in the Makefile? I would restrict Makefile to building the binaries only, and then have all docker packaging etc. be done in the script.

Yes, we could. I've moved all code related to building and uploading multi-arch docker images from Makefile to the scripts.

yurishkuro · 2021-04-25T22:51:29Z

Makefile

+    	$(ANONYMIZER_TAG) \
+		cmd/anonymizer/
+	@echo "Finished building multiarch jaeger-anonymizer =============="
+
 .PHONY: docker-push


we should delete this target, it's not used from GH Actions, and it probably incorrect given these changes.

OK, I've deleted this target.

1. Extracted duplicate codes from build-all-in-one-image.sh and build-upload-docker-images.sh and put them in new scripts. 2. Moved the multi-arch images building and uploading codes from Makefile to scripts. 3. Removed some unused Makefile targets and extra whitespaces. Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-04-27T13:47:59Z

Hi @yurishkuro , thank you so much for your review! I've made changes as per your comments, please take a look when you are available. Thanks again!

Ashmita152

It's looking great. One minor feedback: Do we need to append for-multiarch-image in the script names.

scripts/build-upload-docker-images.sh

Ashmita152 · 2021-04-28T03:32:11Z

scripts/build-upload-docker-images.sh

+		docker buildx build --output "${PUSHTAG}" \
+    	--progress=plain --target release \
+    	--build-arg base_image="localhost:5000/baseimg:1.0.0-alpine-3.12" \
+			--build-arg debug_image="golang:1.15-alpine" \


We don't use this debug_image. We build that too ourselves like baseimg (eg. https://github.com/jaegertracing/jaeger/blob/master/scripts/build-all-in-one-image.sh#L66)

Yes, you're right, we've noticed this. But building debugimg:1.0.0-golang-1.15-alpine needs go-delve which does not have s390x support yet. So at this moment we only make release version of docker images (like all-in-one, jaeger-agent, jaeger-tracegen, etc.) multi-arch and use debug_image="golang:1.15-alpine" instead in build-arg. In fact, the argument debug_image is not used when building release version of images.

The debug version of images (like all-in-one-debug or jaeger-agent-debug ) still use debug_image=localhost/debugimg:1.0.0-golang-1.15-alpine and are amd64 specific.

In fact, the argument debug_image is not used when building release version of images.

If this argument is not used, is it mandatory to pass it ? It just creates confusion, isn't it ?

Yes, since the debug images also use the same Dockerfile, we need to keep this arg and pass it. Although it would create confusion at this moment, once go-delve is supported by s390x, we can use this arg for building multi-arch debug images. Otherwise we need to separate each of these Dockerfiles into two Dockerfiles, one for release target, the other for debug target.

Ashmita152 · 2021-04-28T03:37:51Z

scripts/build-upload-docker-images.sh

+		--platform=${PLATFORMS} \
+    $(echo ${IMAGE_TAGS} | sed "s/JAEGERCOMP/es-index-cleaner/g") \
+		plugin/storage/es
+	docker buildx build --output "${PUSHTAG}" \


Can we make es-index-cleaner, es-rollover, tracegen and anonymizer build in one for loop like you did for jaeger components above.

Yes, we can, I've updated the code to implement this.

Ashmita152 · 2021-04-28T03:57:20Z

scripts/compute-tag-for-multiarch-image.sh

+
+## if we are on a release tag, let's extract the version number
+## the other possible value, currently, is 'master' (or another branch name)
+if [[ $BRANCH == v* ]]; then


I think we should use strict regex match here - something like

$BRANCH =~ ^v[0-9]+\.[0-9]+\.[0-9]+$

Otherwise any PR with branchname starting with letter v will match here

Yes, you're right, I've updated the code to use strict regex match.

1. fix the indentation issues in build-upload-docker-images.sh 2. make other multi-arch images built in one for loop 3. use strict regex to match branch names 4. remove “for-multiarch-image” from script names Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-04-28T19:15:02Z

It's looking great. One minor feedback: Do we need to append for-multiarch-image in the script names.

@Ashmita152 Thank you very much! I've removed for-multiarch-image from script names and made code changes as per your review comments. Please take a look when you are available.

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-05-03T22:28:19Z

@yurishkuro Could you please review this again? Thank you very much!

yurishkuro · 2021-05-04T03:39:04Z

@kun-lu20 could you please describe what kind of testing you did to make sure this all works? Ideally you would run this in your fork with a tag and override the target repos in the registries to your private ones, so that the end result are the images pushed to registries that can be validated.

kun-lu20 · 2021-05-04T15:15:38Z

@yurishkuro Thanks for your response! Yes, each time before I committed the code changes, I would create a testing branch in our fork with the name vM.N.P, and replaced the registry namespace and credentials with my private ones, so that the code changes and GA workflows (including building and pushing the images) could be tested in our fork.
The built images were pushed onto repos on docker.io and quay.io registries under my private namespace kunlu20 (publicly available) and were primitively validated as per the instructions on https://www.jaegertracing.io/docs/1.22/getting-started/ and https://www.jaegertracing.io/docs/1.22/deployment/.

Ashmita152 · 2021-05-04T18:35:27Z

scripts/build-upload-docker-images.sh

+
+# Only push multi-arch images to dockerhub/quay.io for master branch or for release tags vM.N.P
+if [[ "$BRANCH" == "master" || $BRANCH =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+  echo "build multiarch images and upload to dockerhub/quay.io, BRANCH=$BRANCH"


I can see that in some places within this script you are using space while at other places you are using tab for indentation.

I've updated this script to use tab for indentation. Thanks.

Ashmita152 · 2021-05-04T18:36:02Z

scripts/compute-tags.sh

+
+set -exu
+
+###############Compute the tag


can we change this to just single #

Yes, changes done.

Ashmita152 · 2021-05-04T18:36:28Z

scripts/compute-tags.sh

+
+IMAGE_TAGS="${IMAGE_TAGS} --tag docker.io/${SNAPSHOT_TAG} --tag quay.io/${SNAPSHOT_TAG}"
+
+################################


would suggest to remove it.

OK, removed

Ashmita152 · 2021-05-04T18:40:52Z

scripts/build-upload-docker-images.sh

+		else
+			docker_file_arg=""
+		fi
+		if [[ "${component}" =~ ^es-.* ]]; then


Won't this regex match for es-rollover too ? Is that intended ?

This multiple if-else condition is making things confusing. Can we make it cleaner with either switch-case or some other way ?

a simpler approach is to define a function with args like docker_file_arg and then call the function for each component with necessary args

@Ashmita152 Yes, this regex match for both es-index-cleaner and es-rollover , since the value of their dir_arg is the same (plugin/storage/es).
The if-else condition before this is specific for es-rollover because it has an extra --file argument.

I've separated out a function docker_buildx_build() in this script and call it for each component with necessary args. Thanks.

1. Add a function docker_buildx_build() in scripts/ build-upload-docker-images.sh 2. Use tab for indentation in build-upload-docker-images.sh 3. Remove extra '#' from compute-tags.sh Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-06-09T15:07:21Z

Hi @yurishkuro , thanks for your suggestions!

I believe we could use env vars, getops and default values to simplify the argument passing.

Since both the image building and uploading process could be done in one docker buildx build command, we'll still keep them in one script.

Keeping crossdock separate would reduce the number of arguments and if/else branches in the shared script and would not induce much redundancy, so it's definitely worth trying it.

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Create shared script build-upload-a-docker-image.sh, other scripts could call it to build and upload multi-arch docker images. Signed-off-by: Kun-Lu <kun.lu@ibm.com>

yurishkuro · 2021-06-14T16:41:11Z

docker/Makefile

-BASE_IMAGE := localhost/baseimg:$(VERSION)-$(shell echo $(ROOT_IMAGE) | tr : -)
-DEBUG_IMAGE := localhost/debugimg:$(VERSION)-$(shell echo $(GOLANG_IMAGE) | tr : -)
+BASE_IMAGE := localhost:5000/baseimg:$(VERSION)-$(shell echo $(ROOT_IMAGE) | tr : -)
+DEBUG_IMAGE := localhost:5000/debugimg:$(VERSION)-$(shell echo $(GOLANG_IMAGE) | tr : -)


Why do we encode the versions of base images instead of giving them a stable local name? When we include versions we're creating cross-script dependencies which require other scripts to be cognizant of the versions, e.g. base_debug_img_arg="--build-arg base_image=localhost:5000/baseimg:1.0.0-alpine-3.13 --build-arg debug_image=golang:1.15-alpine "

Yes, you're right, I'll change the names of base image and debug image to stable local names. Thanks!

Hi @yurishkuro , I've changed the names of baseimg and debugimg to stable local names. I still set the debug_image arg to golang:1.15-alpine for building multi-arch release version images in build-all-in-one-image.sh and build-upload-docker-images.sh, since the built debugimg doe not support s390x yet (due to lack of go-delve support on s390x). For building amd64 debug version images, the debug_image arg is set to the stable local name.

Hi @yurishkuro , I've made some changes to remove cross-script dependencies of debug_image image names. Please take a look when you are available. Thanks!

docker/Makefile

scripts/build-crossdock.sh

scripts/compute-tags.sh

Change the names of base image and debug image to stable local names and update build-crossdock.sh Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

1. Update docker/debug/Dockerfile to use `if statement` to support s390x 2. Use the stable local name for debugimg across all archs 3. Populate arg `base_debug_img_arg` with the stable local names directly in the shared script `build-upload-a-docker-image.sh` Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-06-23T16:12:57Z

Hi @jpkrohling @yurishkuro , could you please have a look at the recent code changes? I am looking forward to hearing the comments from you. Thank you very much!

jpkrohling

Looks OK to me, had only a couple of comments. But in general, I think we could think a bit about what the ideal world would be in terms of building both the binaries and the images. I feel like there are quite a few things that can be improved.

Makefile

docker/debug/Dockerfile

scripts/build-crossdock.sh

scripts/build-upload-a-docker-image.sh

scripts/build-upload-docker-images.sh

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

jpkrohling

LGTM. I'll give a couple of days for @yurishkuro and @Ashmita152 to review it again, in case they are interested. If I don't hear complaints from them, I'll merge this on Wed.

jpkrohling · 2021-06-30T12:37:20Z

@kun-lu20, could you please rebase this PR? Looks like I can't rebase it myself. Once it's done, I'll merge this.

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 · 2021-06-30T13:15:05Z

@jpkrohling Sure, I've merged the latest updates in branch master into this PR branch. I think you could merge it now. Thanks!

kun-lu20 · 2021-07-01T01:16:37Z

Thanks @yurishkuro @jpkrohling !

kun-lu20 requested a review from a team as a code owner April 21, 2021 16:45

kun-lu20 requested a review from yurishkuro April 21, 2021 16:45

mergify bot assigned jpkrohling Apr 21, 2021

mergify bot requested a review from jpkrohling April 21, 2021 16:46

Add s390x support on multiarch docker images

bae83d3

This change is to add s390x support on multiarch Jaeger docker images Signed-off-by: Kun Lu <kun.lu@ibm.com> Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 force-pushed the enablement_docker_images_on_s390x branch from 93842d0 to bae83d3 Compare April 21, 2021 16:59

Merge branch 'master' into enablement_docker_images_on_s390x

4c6cabb

Signed-off-by: Kun-Lu kun.lu@ibm.com

Ashmita152 reviewed Apr 23, 2021

View reviewed changes

Update tag computation and Dockerfile

c0de24a

Remove chmod from all-in-one Dockerfile Optimize tag computation code Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 force-pushed the enablement_docker_images_on_s390x branch from da1e074 to c0de24a Compare April 23, 2021 20:34

yurishkuro reviewed Apr 25, 2021

View reviewed changes

kun-lu20 requested a review from yurishkuro April 27, 2021 13:43

Ashmita152 reviewed Apr 28, 2021

View reviewed changes

kun-lu20 added 2 commits April 28, 2021 13:54

Optimize the code changes - update

5273aa3

1. fix the indentation issues in build-upload-docker-images.sh 2. make other multi-arch images built in one for loop 3. use strict regex to match branch names 4. remove “for-multiarch-image” from script names Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Merge branch 'master' into enablement_docker_images_on_s390x

ef2e8a2

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Merge branch 'master' into enablement_docker_images_on_s390x

edb391d

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Ashmita152 reviewed May 4, 2021

View reviewed changes

kun-lu20 added 2 commits May 5, 2021 09:03

Optimize the code changes - update

7974892

1. Add a function docker_buildx_build() in scripts/ build-upload-docker-images.sh 2. Use tab for indentation in build-upload-docker-images.sh 3. Remove extra '#' from compute-tags.sh Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Merge branch 'master' into enablement_docker_images_on_s390x

a0362fc

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 added 2 commits June 11, 2021 13:30

Merge branch 'master' into enablement_docker_images_on_s390x

54ba010

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Create a common shared script for docker images

027376a

Create shared script build-upload-a-docker-image.sh, other scripts could call it to build and upload multi-arch docker images. Signed-off-by: Kun-Lu <kun.lu@ibm.com>

yurishkuro reviewed Jun 14, 2021

View reviewed changes

Refactor update

5a6b268

Change the names of base image and debug image to stable local names and update build-crossdock.sh Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 force-pushed the enablement_docker_images_on_s390x branch from 03c360e to 5a6b268 Compare June 14, 2021 20:36

kun-lu20 requested a review from yurishkuro June 14, 2021 20:49

kun-lu20 added 4 commits June 14, 2021 17:02

Merge branch 'master' into enablement_docker_images_on_s390x

2960e88

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Merge branch 'master' into enablement_docker_images_on_s390x

2acd49f

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Merge branch 'master' into enablement_docker_images_on_s390x

4e4fcbb

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

jpkrohling reviewed Jun 24, 2021

View reviewed changes

Makefile Show resolved Hide resolved

docker/debug/Dockerfile Show resolved Hide resolved

scripts/build-crossdock.sh Show resolved Hide resolved

scripts/build-upload-a-docker-image.sh Outdated Show resolved Hide resolved

scripts/build-upload-docker-images.sh Show resolved Hide resolved

kun-lu20 added 2 commits June 24, 2021 15:47

Merge branch 'master' into enablement_docker_images_on_s390x

878e983

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Fix the indentation issue

18aad23

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

kun-lu20 requested a review from jpkrohling June 24, 2021 19:57

kun-lu20 added 2 commits June 25, 2021 15:00

Merge branch 'master' into enablement_docker_images_on_s390x

56f3e4f

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

Add comment in debug Dockerfile

c492340

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

jpkrohling approved these changes Jun 28, 2021

View reviewed changes

kun-lu20 mentioned this pull request Jun 28, 2021

Rethink our CI to avoid building all the docker images every time #3113

Open

GaruGaru mentioned this pull request Jun 29, 2021

added multi arch images builds support for make file #2611

Closed

2 tasks

yurishkuro approved these changes Jun 29, 2021

View reviewed changes

Merge branch 'master' into enablement_docker_images_on_s390x

a62c82f

Signed-off-by: Kun-Lu <kun.lu@ibm.com>

yurishkuro enabled auto-merge (squash) June 30, 2021 14:26

yurishkuro merged commit 78d2f7d into jaegertracing:master Jun 30, 2021

yurishkuro mentioned this pull request Jul 1, 2021

arm64 support jaegertracing/jaeger-operator#1086

Closed

Ashmita152 mentioned this pull request Aug 16, 2021

Refactor CI jobs to use goreleaser #3209

Closed


		IMAGE_TAGS="${IMAGE_TAGS} --tag docker.io/${SNAPSHOT_TAG} --tag quay.io/${SNAPSHOT_TAG}"

		################################

Add s390x support on multiarch docker images #2948

Add s390x support on multiarch docker images #2948

Conversation

kun-lu20 commented Apr 21, 2021 • edited by yurishkuro Loading

Which problem is this PR solving?

Short description of the changes

codecov bot commented Apr 21, 2021 • edited Loading

Codecov Report

Ashmita152 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 commented Apr 23, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 commented Apr 27, 2021

Ashmita152 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 Apr 28, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 commented Apr 28, 2021

kun-lu20 commented May 3, 2021

yurishkuro commented May 4, 2021

kun-lu20 commented May 4, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Ashmita152 May 4, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 May 4, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 commented Jun 9, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 Jun 14, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kun-lu20 commented Jun 23, 2021

jpkrohling left a comment

Choose a reason for hiding this comment

jpkrohling left a comment

Choose a reason for hiding this comment

jpkrohling commented Jun 30, 2021

kun-lu20 commented Jun 30, 2021

kun-lu20 commented Jul 1, 2021

kun-lu20 commented Apr 21, 2021 •

edited by yurishkuro

Loading

codecov bot commented Apr 21, 2021 •

edited

Loading

kun-lu20 Apr 28, 2021 •

edited

Loading

kun-lu20 commented May 4, 2021 •

edited

Loading

Ashmita152 May 4, 2021 •

edited

Loading

kun-lu20 May 4, 2021 •

edited

Loading

kun-lu20 Jun 14, 2021 •

edited

Loading