| 1 |
[[s-kubernetes-upgrade]] |
| 2 |
=== Upgrading a LINSTOR Deployment on Kubernetes |
| 3 |
|
| 4 |
A LINSTOR Deployment on Kubernets can be upgraded to a new release using Helm. |
| 5 |
|
| 6 |
Before upgrading to a new release, you should ensure you have an up-to-date backup of the LINSTOR database. |
| 7 |
If you are using the Etcd database packaged in the LINSTOR Chart, see <<s-kubernetes-etcd-backup,here>> |
| 8 |
|
| 9 |
IMPORTANT: Upgrades using the LINSTOR Etcd deployment require etcd to use persistent storage. Only follow these steps if |
| 10 |
Etcd was deployed using `etcd.persistentVolume.enabled=true` |
| 11 |
|
| 12 |
Upgrades will update to new versions of the following components: |
| 13 |
|
| 14 |
* LINSTOR operator deployment |
| 15 |
* LINSTOR Controller |
| 16 |
* LINSTOR Satellite |
| 17 |
* LINSTOR CSI Driver |
| 18 |
* Etcd |
| 19 |
* Stork |
| 20 |
|
| 21 |
Some versions require special steps, please take a look <<s-kubernetes-upgrade-version,here>> |
| 22 |
The main command to upgrade to a new LINSTOR operator version is: |
| 23 |
|
| 24 |
---- |
| 25 |
helm repo update |
| 26 |
helm upgrade linstor-op linstor/linstor |
| 27 |
---- |
| 28 |
|
| 29 |
If you used any customizations on the initial install, pass the same options to `helm upgrade`. The options currently |
| 30 |
in use can be retrieved from Helm. |
| 31 |
|
| 32 |
---- |
| 33 |
# Retrieve the currently set options |
| 34 |
$ helm get values linstor-op |
| 35 |
USER-SUPPLIED VALUES: |
| 36 |
USER-SUPPLIED VALUES: null |
| 37 |
drbdRepoCred: drbdiocred |
| 38 |
operator: |
| 39 |
satelliteSet: |
| 40 |
kernelModuleInjectionImage: drbd.io/drbd9-rhel8:v9.0.28 |
| 41 |
storagePools: |
| 42 |
lvmThinPools: |
| 43 |
- devicePaths: |
| 44 |
- /dev/vdb |
| 45 |
name: thinpool |
| 46 |
thinVolume: thinpool |
| 47 |
volumeGroup: "" |
| 48 |
# Save current options |
| 49 |
$ helm get values linstor-op > orig.yaml |
| 50 |
# modify values here as needed. for example selecting a newer DRBD image |
| 51 |
$ vim orig.yaml |
| 52 |
# Start the upgrade |
| 53 |
$ helm upgrade linstor-op linstor/linstor -f orig.yaml |
| 54 |
---- |
| 55 |
|
| 56 |
This triggers the rollout of new pods. After a short wait, all pods should be running and ready. |
| 57 |
Check that no errors are listed in the status section of LinstorControllers, LinstorSatelliteSets and LinstorCSIDrivers. |
| 58 |
|
| 59 |
IMPORTANT: During the upgrade process, provisioning of volumes and attach/detach operations might not work. Existing |
| 60 |
volumes and volumes already in use by a pod will continue to work without interruption. |
| 61 |
|
| 62 |
[[s-kubernetes-upgrade-k8s-backend]] |
| 63 |
==== Upgrading a LINSTOR Cluster with Kubernetes backend |
| 64 |
|
| 65 |
Using the Kubernetes API as LINSTOR backend is quite new: there might be situations in which LINSTOR fails to upgrade |
| 66 |
resources, leaving the LINSTOR cluster in an unusable state. For this reason, we strongly recommend doing a manual |
| 67 |
backup of all LINSTOR resources before upgrading the cluster. In a future release, this step will be managed by the |
| 68 |
operator. |
| 69 |
|
| 70 |
1. Stop the current controller: |
| 71 |
+ |
| 72 |
---- |
| 73 |
$ kubectl patch linstorcontroller linstor-cs '{"spec":{"replicas": 0}}' |
| 74 |
$ kubectl rollout status --watch deployment/linstor-cs-controller |
| 75 |
---- |
| 76 |
2. The following command will create a file `crds.yaml`, which stores the current state of all LINSTOR Custom Resource |
| 77 |
Definitions: |
| 78 |
+ |
| 79 |
---- |
| 80 |
$ kubectl get crds | grep -o ".*.internal.linstor.linbit.com" | xargs kubectl get crds -oyaml > crds.yaml |
| 81 |
---- |
| 82 |
+ |
| 83 |
3. In addition to the definitions, the actual resources must be be backed up as well: |
| 84 |
+ |
| 85 |
---- |
| 86 |
$ kubectl get crds | grep -o ".*.internal.linstor.linbit.com" | xargs -i{} sh -c "kubectl get {} -oyaml > {}.yaml" |
| 87 |
---- |
| 88 |
4. Run the chart upgrade using `--set IHaveBackedUpAllMyLinstorResources=true` to acknowledge you have executed the |
| 89 |
above steps. |
| 90 |
|
| 91 |
[[s-kubernetes-upgrade-version]] |
| 92 |
==== Upgrade instructions for specific versions |
| 93 |
|
| 94 |
Some versions require special steps, see below. |
| 95 |
|
| 96 |
===== Upgrade to v1.7 |
| 97 |
|
| 98 |
No additional steps necessary. |
| 99 |
|
| 100 |
===== Upgrade to v1.6 |
| 101 |
|
| 102 |
This versions introduces a new option to support Kubernetes distributions which use different state directories than the |
| 103 |
default of `/var/lib/kubelet`. A notable example is microk8s, which uses `/var/snap/microk8s/common/var/lib/kubelet`. |
| 104 |
To support this, a small addition to the `LinstorCSIDriver` CRD was necessary. As Helm does not upgrade CRDs on chart |
| 105 |
upgrade, instead please run: |
| 106 |
|
| 107 |
---- |
| 108 |
$ helm repo update |
| 109 |
$ helm pull linstor/linstor --untar |
| 110 |
$ kubectl replace -f linstor/crds/ |
| 111 |
customresourcedefinition.apiextensions.k8s.io/linstorcontrollers.linstor.linbit.com replaced |
| 112 |
customresourcedefinition.apiextensions.k8s.io/linstorcsidrivers.linstor.linbit.com replaced |
| 113 |
customresourcedefinition.apiextensions.k8s.io/linstorsatellitesets.linstor.linbit.com replaced |
| 114 |
---- |
| 115 |
|
| 116 |
If you do not apply the new CRDs, you will get errors like the following: |
| 117 |
|
| 118 |
---- |
| 119 |
Error: UPGRADE FAILED: error validating "": error validating data: ValidationError(LinstorCSIDriver.spec): unknown field "kubeletPath" in com.linbit.linstor.v1.LinstorCSIDriver.spec |
| 120 |
---- |
| 121 |
|
| 122 |
If you previously used the included snapshot-controller to process `VolumeSnapshot` resources, you should replace it |
| 123 |
with the new charts provided by the Piraeus project. The <<s-kubernetes-add-snaphot-support,section on snapshots>> has |
| 124 |
been updated to include instructions on how you can add the snapshot-controller to your cluster. |
| 125 |
|
| 126 |
===== Upgrade to v1.5 |
| 127 |
|
| 128 |
This version introduces a <<s-kubernetes-monitoring,monitoring>> component for DRBD resources. This requires a new image |
| 129 |
and a replacement of the existing `LinstorSatelliteSet` CRD. Helm does not upgrade the CRDs on a chart upgrade, |
| 130 |
instead please run: |
| 131 |
|
| 132 |
---- |
| 133 |
$ helm repo update |
| 134 |
$ helm pull linstor/linstor --untar |
| 135 |
$ kubectl replace -f linstor/crds/ |
| 136 |
customresourcedefinition.apiextensions.k8s.io/linstorcontrollers.linstor.linbit.com replaced |
| 137 |
customresourcedefinition.apiextensions.k8s.io/linstorcsidrivers.linstor.linbit.com replaced |
| 138 |
customresourcedefinition.apiextensions.k8s.io/linstorsatellitesets.linstor.linbit.com replaced |
| 139 |
---- |
| 140 |
|
| 141 |
If you do not plan to use the provided <<s-kubernetes-monitoring,monitoring>> you still need to apply the above steps, |
| 142 |
otherwise you will get an errors like the following |
| 143 |
|
| 144 |
---- |
| 145 |
Error: UPGRADE FAILED: error validating "": error validating data: ValidationError(LinstorSatelliteSet.spec): unknown field "monitoringImage" in com.linbit.linstor.v1.LinstorSatelliteSet.spec |
| 146 |
---- |
| 147 |
|
| 148 |
NOTE: Some Helm versions fail to set the monitoring image even after replacing the CRDs. In that case, the in-cluster |
| 149 |
LinstorSatelliteSet will show an empty `monitoringImage` value. Edit the resource using |
| 150 |
`kubectl edit linstorsatellitesets` and set the value to `drbd.io/drbd-reactor:v0.3.0` to enable monitoring. |
| 151 |
|
| 152 |
===== Upgrade to v1.4 |
| 153 |
|
| 154 |
This version introduces a new default version for the Etcd image, so take extra care that Etcd is using |
| 155 |
persistent storage. *Upgrading the Etcd image without persistent storage will corrupt the cluster*. |
| 156 |
|
| 157 |
If you are upgrading an existing cluster without making use of new Helm options, no additional steps are necessary. |
| 158 |
|
| 159 |
If you plan to use the newly introduced `additionalProperties` and `additionalEnv` settings, you have to replace |
| 160 |
the installed CustomResourceDefinitions with newer versions. Helm does not upgrade the CRDs on a chart upgrade |
| 161 |
|
| 162 |
---- |
| 163 |
$ helm pull linstor/linstor --untar |
| 164 |
$ kubectl replace -f linstor/crds/ |
| 165 |
customresourcedefinition.apiextensions.k8s.io/linstorcontrollers.linstor.linbit.com replaced |
| 166 |
customresourcedefinition.apiextensions.k8s.io/linstorcsidrivers.linstor.linbit.com replaced |
| 167 |
customresourcedefinition.apiextensions.k8s.io/linstorsatellitesets.linstor.linbit.com replaced |
| 168 |
---- |
| 169 |
|
| 170 |
===== Upgrade to v1.3 |
| 171 |
|
| 172 |
No additional steps necessary. |
| 173 |
|
| 174 |
===== Upgrade to v1.2 |
| 175 |
|
| 176 |
LINSTOR operator v1.2 is supported on Kubernetes 1.17+. If you are using an older Kubernetes distribution, you may need |
| 177 |
to change the default settings, for example [the CSI provisioner](https://kubernetes-csi.github.io/docs/external-provisioner.html). |
| 178 |
|
| 179 |
There is a known issue when updating the CSI components: the pods will not be updated to the newest image and the |
| 180 |
`errors` section of the LinstorCSIDrivers resource shows an error updating the DaemonSet. In this case, manually |
| 181 |
delete `deployment/linstor-op-csi-controller` and `daemonset/linstor-op-csi-node`. They will be re-created by the |
| 182 |
operator. |
| 183 |
|
| 184 |
[[s-kubernetes-etcd-backup]] |
| 185 |
==== Creating Etcd Backups |
| 186 |
|
| 187 |
To create a backup of the Etcd database and store it on your control host, run: |
| 188 |
|
| 189 |
[source] |
| 190 |
---- |
| 191 |
kubectl exec linstor-op-etcd-0 -- etcdctl snapshot save /tmp/save.db |
| 192 |
kubectl cp linstor-op-etcd-0:/tmp/save.db save.db |
| 193 |
---- |
| 194 |
|
| 195 |
These commands will create a file `save.db` on the machine you are running `kubectl` from. |