1.12.0+ Upgrade Notice
Review and apply version-specific changes in Gloo Edge 1.12.x and Gloo Edge Enterprise 1.12.x before upgrading your Gloo Edge installation.
New CRDs are automatically applied to your cluster when performing a
helm install operation.
However, they are not applied when performing an
helm upgrade operation. This is a deliberate design choice on the part of the
Helm maintainers, given the risk associated with changing CRDs. Given this limitation, you must apply new CRDs to the cluster before upgrading.
To reduce cross-pod communication, the gateway functionality is removed from a standalone pod, and is now included in the Gloo Edge control plane pod. Specifically, the validation webhook is moved to the
gloo pod, which reduces the number of gRPC calls required for validation.
The following changes apply to your Gloo Edge installation in 1.12.x:
- Installed pods: The
gatewaypod is removed from the set of installed components.
- Installation settings:
- The sections of the Helm chart for the gateway deployment and service accounts are removed.
- Because the
gatewaypod is no longer installed, specify any environment variables for the gateway on the
gloodeployment instead. For example, in your Helm values file, you might set the
PROXY_STATUS_MAX_SIZE_BYTESenvironment variable to truncate the length of the proxy status field:
gloo: deployment: customEnv: - name: "PROXY_STATUS_MAX_SIZE_BYTES" value: "100"
- To improve performance, the proxy CRD is not written to etcd by default and is instead kept in memory. If you must allow external tools and other pods to read the proxy CR contents, you can persist them to etcd by setting the
truein your Helm values file. Important: If you use the Kubernetes ExternalDNS functionality, you must set the
- Proxy CR contents: To view the contents of a proxy CR, you can continue to use the
glooctl get proxy <proxy> -o yamlcommand. However, because the proxy CRD is no longer written to etcd by default, you cannot use
kubectlcommands to check proxy CR contents.
- Snapshot metrics: The
gatewaysnapshot is removed. If you use metrics from the gateway snapshot, those metrics are moved to the
gloosnapshot; for example,
In Gloo Edge v1.12.0 / Gloo Edge Enterprise v1.12.0 and later, the default Gateways, ExtAuth Upstreams, and RateLimit Upstreams are installed via a Job during the Helm installation and upgrade process, instead of being included directly in the Helm chart. After you upgrade to these versions, those resources no longer have Helm annotations (e.g.
meta.helm.sh/release-namespace). If you need to roll back to an earlier Gloo Edge version, you must manually add the Helm release annotations back to these resources before performing the rollback.
Do not manually edit (e.g.
kubectl apply or
kubectl edit) the Gloo Edge custom resources (e.g. Gateways, Upstreams) that are installed with the Helm chart, as this might cause errors in future upgrades.
If you already manually edited a resource, verify before you upgrade that the resource’s
metadata.annotations.kubectl.kubernetes.io/last-applied-configuration annotation does not contain a
resourceVersion field. This field can cause conflicts during the upgrade process.
If the field exists, you can remove it in either of the following ways:
kubectl editthe resource and manually delete the
resourceVersionfield from the
- remove the
metadata.annotations.kubectl.kubernetes.io/last-applied-configurationannotation altogether if not needed, e.g.
kubectl annotate gateway -n gloo-system gateway-proxy kubectl.kubernetes.io/last-applied-configuration-
New and updated CRDs
- In the ExtAuth CR, the
spec.configs.oauth2.oidcAuthorizationCode.headerssection is updated to include the
useBearerSchemaForAuthorizationsetting to add the “Bearer” prefix to the upstream access token header value (
"Authorization: Bearer <access_token>"). In Gloo Edge 1.12, this setting defaults to false. In Gloo Edge 1.13, this setting will default to true. For more information, see the ExtAuth reference documentation.
Deprecated CRDs None
Removed CRDs None
To upgrade Gloo Edge:
Follow steps 1 - 2 in Upgrade Steps to prepare for upgrading, and to upgrade
Apply the new and updated CRDs. Replace the version with the specific patch version that you are upgrading to.
helm repo update helm pull gloo/gloo --version 1.12.0 --untar kubectl apply -f gloo/crds
helm repo update helm pull glooe/gloo-ee --version 1.12.0 --untar kubectl apply -f gloo-ee/charts/gloo/crds # If Gloo Federation is enabled kubectl apply -f gloo-ee/charts/gloo-fed/crds
Prepare for any updates you want to make related to the gateway architectural changes. For example, you might want to create a Helm values file to pass gateway environment variables to the
gloopod during the Helm upgrade, or update any systems that watch for snapshot metrics.
Continue to upgrade the Gloo Edge server components via Helm.
If you use the
spec.configs.oauth2.oidcAuthorizationCode.headers.accessTokenHeader: "Authorization"setting in any
AuthConfigresources, and you do not want to add the “Bearer” prefix to the header, set
false. If you do not set this setting, it will default to true in Gloo Edge 1.13.