I'm trying to get list of all the tags in my private GCR repository. I could do that using "gcloud container images list-tags" command as follows:
gcloud container images list-tags gcr.io/project-id/REPONAME
DIGEST TAGS TIMESTAMP
6b5727be962a 0.0.4,latest 2020-06-25T14:14:48
4b8c3f9c6ab7 0.0.3 2020-06-22T08:56:01
However I need the list to be flatten so that i can get tags "0.0.4" and "latest" in separate rows. I tried following command.
gcloud container images list-tags gcr.io/project-id/REPONAME --flatten='[].tags'
This gave me output which is to my surprise repeating "latest" tag but ommiting "0.0.4"
DIGEST TAGS TIMESTAMP
6b5727be962a latest 2020-06-25T14:14:48
6b5727be962a latest 2020-06-25T14:14:48
4b8c3f9c6ab7 0.0.3 2020-06-22T08:56:01
What am I doing wrong, and how can I fix this?
I am able to repro your observation and think it's a bug.
The --flatten appears to correctly enumerate tags but incorrectly returns the last value in the list as each entry's value.
In my case, if the tags are v1,v2,v3, I get:
gcloud container images list-tags gcr.io/${PROJECT}/${IMAGE} \
--flatten="[].tags[]" \
--format="value(tags)" \
--filter="digest=${DIGEST}"
v3
v3
v3
I recommend you file a bug on Google's Issue Tracker for Cloud SDK
If you have jq, perhaps:
gcloud container images list-tags gcr.io/${PROJECT}/${IMAGE} \
--format=json |\
jq -r '.[] | .digest as $D | .timestamp.datetime as $T | .tags[]| {"digest":$D,"tag":.,"timestamp":$T}'
Or:
gcloud container images list-tags gcr.io/${PROJECT}/${IMAGE} \
--format=json |\
jq -r '.[] | .digest as $D | .timestamp.datetime as $T | .tags[]| [$D,.,$T] | @csv'