Signed-off-by: Parul Singh [email protected]

Which issue(s) this PR addresses: For podman deployment:

• systemd unit file for starting and managing microshift-containerized.

Closes #434, #433, #432

Cluster DNS is set to the 10th IP of Service CIDR

Signed-off-by: Vu Dinh [email protected]

Which issue(s) this PR addresses:

Closes #

klog is a singleton library, and since we use that library from all our services in a single process setting different log files won't work, and will syphon all logs into the last log file we add.

Keep the simple strategy of letting all output to stderr for now.

Related-Issue: #493

Signed-off-by: Miguel Angel Ajo [email protected]

Separate component-specific config and microshift-specific config into 2 sections within top-level config. The components config is modelled after OpenShift config v1 API while microshift config contains fields that are specific to microshift usage.

Signed-off-by: Vu Dinh [email protected]

Which issue(s) this PR addresses:

Closes #

Closes: USHIFT-233

This is related to https://github.com/openshift/enhancements/pull/1203, but we can go ahead and take it now to make less work to update test suite changes later.

Signed-off-by: Parul [email protected]

Which issue(s) this PR addresses:

Closes https://github.com/redhat-et/microshift/issues/134

Which issue(s) this PR addresses:

Closes USHIFT-227

This PR carry the following items:

• Enabling Cluster Policy Controller
• Disabling resource-quota and cluster-quota-reconciliation controllers.
• Creating of openshift-kube-controller-manager namespace (it's where the CreatedSCCRanges events happen)
• Applying csr-approver and namespace-security cluster roles and cluster role bindings. (required by the CPC)

What would you like to be added:

Monitoring Platform/Cluster Monitoring Nodes Monitoring Pods Containers status Pods per Node Services Health Resources Utilization (CPU/Memory/Network) etc.

Why is this needed:

Give end-user visibility, on the state and health of their applications and solutions Very important for the mission-critical applications

Allow MicroShift to join new worker nodes, according to design here #498 (see individual commits for review),

• 4523987f Add flags to allow TLS bootstrapping of nodes
• 9d8e2ade Add bootstrap module and generate token file
• 26f04a33 Add ClusterRoleBinding for bootstrap process
• 1103d731 Generate bootstrap kubeconfig
• 1fd2a3c9 Allow MicroShift to start node role standalone
• 6c84f168 Apply CRB for bootstraping nodes
• 1366626f Use bootstrap kubeconfig for kube-proxy
• c2eacfd6 Use netcgo insted of netgo
• 40bff393 Add vagrant env to test/devel/debug multi-worker

Related PRs: #499 , #500

This commit only describes the addition of new compute nodes to an existing MicroShift cluster. Highly available control plane will be described in later PRs.

Signed-off-by: Ricardo Noriega [email protected]

This Enhacement proposal addresses part of the #460 epic.

$ cat<<EOF | oc apply --server-side -f-
apiVersion: route.openshift.io/v1
kind: Route
metadata:
  name: hello-microshift
spec:
  to:
    kind: Service
    name: hello-microshift
EOF

route.route.openshift.io/hello-microshift serverside-applied

$ oc get route hello-microshift -o yaml

apiVersion: route.openshift.io/v1
kind: Route
metadata:
  annotations:
    openshift.io/host.generated: "true"
  creationTimestamp: "2022-11-11T23:53:33Z"
  generation: 1
  name: hello-microshift
  namespace: default
  resourceVersion: "2659"
  uid: cd35cd20-b3fd-4d50-9912-f34b3935acfd
spec:
  host: hello-microshift-default.cluster.local
  to:
    kind: Service
    name: hello-microshift
  wildcardPolicy: None

$ cat<<EOF | oc apply --server-side -f-
apiVersion: route.openshift.io/v1
kind: Route
metadata:
  name: hello-microshift
spec:
  to:
    kind: Service
    name: hello-microshift
  wildcardPolicy: ""
EOF

The Route "hello-microshift" is invalid: spec.wildcardPolicy: Invalid value: "": field is immutable

As of now assets/release/release-aarch64.json file have arch info as part of base section "base": "4.12.0-0.nightly-arm64-2023-01-03-161334" but it is not present for assets/release/release-x86_64.json. This PR puts the arch info in base same as aarch64 asset file.

Wraps rebase.sh with extra goodies for better interactions with PRs. Checks rebase.sh exit code, so case of error the logs is committed along of any other staging changes, PR's title will get *FAILURE* prefix and PR's description will have big (h1) message to check committed rebase.sh log. If PR exists for the branch (which name is amd64's release tag), PR desc will be updated. PR description now includes: amd64 and arm64 release tags, and (deduced) prow job link.

Obsoletes create_pr.py

ovn databases (nb/sb) were stored in containers and recreated every time when ovnkube-master container restarts. This commit saves container databases to host directory (/etc/ovn/) which avoids recreation of ovn databases after ovnkube-master restarts. It optimizes around 0~10M memory for each database container from a snapshot metrics with metrics-server. Also add a command in cleanup script to remove the ovn databases.

Related-Issue: https://issues.redhat.com/browse/NP-648

What happened?

trying to install the microshift on my lab and getting error "dracut-initqueue timeout - starting timeout scripts". Following guide https://github.com/openshift/microshift/blob/main/docs/getting_started.md

Just before the dracut timeout the lines are shown.

Tech Preview: NMVe/TCP may not be fully supported.

Using the latest RHEL 8.7 image. Most likely issue with the RHEL.

What did you expect to happen?

Normal installation.

How to reproduce it (as minimally and precisely as possible)?

1. '...'
2. '...'

Anything else we need to know?

Environment

• MicroShift version (use microshift version):
• Hardware configuration:
• OS (e.g: cat /etc/os-release): RHEL 8.7 on Fedora 35
• Kernel (e.g. uname -a): Linux 6.0.12-100.fc35.x86_64

Relevant logs