Release period: 2021-08-11 to 2021-08-18
This release includes the following issues:
- Add PersistentVolumeClaims to OpenShift/Kubernetes metering
- Provide OpenShift/Kubernetes localIds in meshObject API
- Auto-Approval of quota requests up to defined threshold
- Fix OpenShift platform type in meshMetering
- Temporary removal of Platform Access chart on landing page
- Flexible OpenShift Pod CPU product catalog units
- Replicator downgrades its own AWS permissions
- Import of user groups without members
Add PersistentVolumeClaims to OpenShift/Kubernetes metering
Audience: Customer, Operator
PersistentVolumeClaims in OpenShift and Kubernetes can now be collected and charged by meshMetering. A price can be defined per capacity of the PersistentVolumeClaim. Different prices can be defined according to the storage class being used and the status of the volume claim.
Provide OpenShift/Kubernetes localIds in meshObject API
OpenShift and Kubernetes tenants did not provide a localId via the meshObject API. These localIds are provided correctly now.
Auto-Approval of quota requests up to defined threshold
Audience: Customer, Partner, Operator
It is now possible to define an auto-approval threshold per Platform Quota Definition. If a quota request is created where all quotas are below the according defined threshold, the request will be auto-approved. This will reduce the load of platform operators regarding approval of quota requests. It will also reduce the time for customers to wait for an approval to zero as long as their request is below the defined threshold. The auto-approval will also be applied when only reducing quotas.
Fix OpenShift platform type in meshMetering
OpenShift metering data was assigned to platform type Kubernetes. This didn't actually break anything, but it e.g. showed the Kuberentes logo instead of the OpenShift logo in Usage Reports and Chargeback Statements. Additionally type Kubernetes had to be defined for OpenShift platforms in the Price Catalog. This has been fixed now and OpenShift is handled in meshMetering with the correct platform type now.
Temporary removal of Platform Access chart on landing page
meshStack displays a Platform Access chart in the bottom of the meshcloud landing page. We have identified an issue in a third party solution which is being used to generate this chart (for details see https://github.com/grafana/grafana-image-renderer/issues/269). This issue has the possibility to harm other monitoring components in a meshstack. As a result, we are temporarily removing the component to protect the remaining monitoring components. We are actively looking both with the authors of the third-party component as well as internally for a solution.
Flexible OpenShift Pod CPU product catalog units
Operators can now configure product catalog entries for OpenShift Pod CPU usage with prefixed units. Previously products could only be specified on "milli CPU" in the units of mCPU. Operators can now define prices with any SI-prefix, e.g. "mCPU", "CPU" or "kCPU". This allows presenting OpenShift pod consumption in human-readable and intuitive units.
Replicator downgrades its own AWS permissions
After the replicator provisions a new AWS account, it initially has administrator access in this new account. Previously, some of you would have implemented your own mechanism via CloudFormation or Lambda to downgrade this role to have the minimum required access. From this release onwards, the replicator is capable of downgrading its own permissions. If you already have custom mechanisms to downgrade this role, please remove those mechanisms before deploying this release.
Import of user groups without members
Customer user group having no members can now be successfully imported via the identity connector.