Skip to content

Fix couple multicluster flaky tests #562

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Oct 29, 2025
Merged

Fix couple multicluster flaky tests #562

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion .evergreen-tasks.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1116,7 +1116,6 @@ tasks:

- name: e2e_multi_cluster_validation
tags: [ "patch-run" ]
exec_timeout_secs: 1000
Copy link
Collaborator Author

@MaciejKaras MaciejKaras Oct 29, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this is not needed and prevents us from collecting test data

commands:
- func: e2e_test

Expand Down
11 changes: 2 additions & 9 deletions ...mongodb-kubernetes-tests/tests/multicluster_appdb/multicluster_appdb_disaster_recovery.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -152,11 +152,12 @@ def test_delete_om_and_appdb_statefulset_in_failed_cluster(
# delete OM to simulate losing Ops Manager application
# this is only for testing unavailability of the OM application, it's not testing losing OM cluster
# we don't delete here any additional resources (secrets, configmaps) that are required for a proper OM recovery testing
# it will be immediately recreated by the operator, so we cannot check if it was deleted
delete_statefulset(
ops_manager.namespace,
ops_manager.name,
propagation_policy="Background",
api_client=central_cluster_client,
api_client=get_member_cluster_api_client(OM_MEMBER_CLUSTER_NAME),
)
except kubernetes.client.ApiException as e:
if e.status != 404:
Expand Down Expand Up @@ -184,14 +185,6 @@ def statefulset_is_deleted(namespace: str, name: str, api_client=Optional[kubern
else:
raise e

run_periodically(
Copy link
Collaborator Author

@MaciejKaras MaciejKaras Oct 29, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is racy, because stateful set is immediately recreated by operator

lambda: statefulset_is_deleted(
ops_manager.namespace,
ops_manager.name,
api_client=get_member_cluster_api_client(OM_MEMBER_CLUSTER_NAME),
),
timeout=120,
)
run_periodically(
lambda: statefulset_is_deleted(
ops_manager.namespace,
Expand Down
33 changes: 12 additions & 21 deletions scripts/funcs/multicluster
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -128,26 +128,6 @@ EOF
sleep 1

local service_account_name="operator-tests-multi-cluster-service-account"

Copy link
Collaborator Author

@MaciejKaras MaciejKaras Oct 29, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

duplicated code that is already present jus below. Creating Multi Cluster configuration secret shows twice in the logs as well

local secret_name
secret_name="$(kubectl --context "${CENTRAL_CLUSTER}" get secret -n "${NAMESPACE}" | { grep "${service_account_name}" || test $? = 1; } | awk '{ print $1 }')"
if [[ "${secret_name}" == "" ]]; then
secret_name="${service_account_name}-token-secret"
create_service_account_token_secret "${CENTRAL_CLUSTER}" "${service_account_name}" "${secret_name}"
fi

local central_cluster_token
central_cluster_token="$(kubectl --context "${CENTRAL_CLUSTER}" get secret "${secret_name}" -o jsonpath='{ .data.token}' -n "${NAMESPACE}" | base64 -d)"
echo "Creating Multi Cluster configuration secret"

configuration_params=(
"--from-literal=central_cluster=${CENTRAL_CLUSTER}"
)

configuration_params+=(
"--from-literal=${CENTRAL_CLUSTER}=${central_cluster_token}"
)

local secret_name
secret_name="$(kubectl --context "${CENTRAL_CLUSTER}" get secret -n "${NAMESPACE}" | { grep "${service_account_name}" || test $? = 1; } | awk '{ print $1 }')"
if [[ "${secret_name}" == "" ]]; then
Expand Down Expand Up @@ -175,7 +155,18 @@ EOF
create_service_account_token_secret "${member_cluster}" "${service_account_name}" "${secret_name}"
fi

member_cluster_token="$(kubectl --context "${member_cluster}" get secret "${secret_name}" -o jsonpath='{ .data.token}' -n "${NAMESPACE}" | base64 -d)"
# Retry up to 10 times if .data.token is not yet populated
Copy link
Collaborator Author

@MaciejKaras MaciejKaras Oct 29, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

secrets .data.token is populated asynchronously after the secret is created. We need to wait for it become available, otherwise the member_cluster_token will be empty

for _ in {1..10}; do
member_cluster_token="$(kubectl --context "${member_cluster}" get secret "${secret_name}" -o jsonpath='{ .data.token }' -n "${NAMESPACE}" | base64 -d)"
if [[ -n "${member_cluster_token}" ]]; then
break
fi
sleep 1
done
if [[ -z "${member_cluster_token}" ]]; then
echo "Error: .data.token not populated for secret ${secret_name} in cluster ${member_cluster}"
exit 1
fi
# for 2 cluster tests central cluster is the first member, so we cannot add this as it will result in duplicate key and error in create secret
if [[ "${member_cluster}" != "${CENTRAL_CLUSTER}" ]]; then
configuration_params+=(
Expand Down
Loading