📝 helm: improve upgrade doc

Author: jfbus

docs/README.md

@@ -141,17 +141,15 @@ Common issues and diagnostics are covered in **[Troubleshooting](./troubleshooti
 
 ### Upgrading from v0.x to v1.0.0
 
-Follow the **[Migration Process](./migration.md)**.
+Follow the [migration guide](./migration.md).
 
 ### Upgrading from v1.6 to v1.7
 
 `maxBsuVolumes` is now computed automatically at driver startup. Manual configuration is usually unnecessary, even when multiple BSU volumes are mounted by the OS.
 
 ### Upgrading to Helm chart v2
 
-Many variables have been renamed.
-
-Please refer to the [upgrade guide](../helm/osc-bsu-csi-driver/migration.md).
+Most values have been renamed, and need to be set again using the new names during the upgrade. Please refer to the [upgrade guide](../helm/osc-bsu-csi-driver/migration.md).
 
 ---
 

helm/osc-bsu-csi-driver/migration.md

@@ -1,29 +1,29 @@
 # Migration from v1 chart to v2 chart
 
-Most variables have been renamed, and `helm upgrade` will fail if called with v1 values.
+Most variables have been renamed, and `helm upgrade` will fail if you reuse v1 values.
 
 You can call `helm get values osc-bsu-csi-driver --namespace kube-system` to display configured values.
-And you will need to reset existing values while upgrading by using `helm upgrade --reset-values`.
+You will need to set each value again using the new value names.
 
 For example, if you installed the CSI driver using:
 
 ```shell
 helm install [...] \
     --set enableVolumeSnapshot=true \
-    --set region=$OSC_REGION
+    --set region=$OSC_REGION \
+    --set verbosity=5
 ```
 
 You will need to upgrade with:
 
 ```shell
-helm upgrade [...] --reset-values \
+helm upgrade [...] \
     --set driver.enableVolumeSnapshot=true \
-    --set cloud.region=$OSC_REGION
+    --set cloud.region=$OSC_REGION \
+    --set logs.verbosity=5
 ```
 
-## Renamed values
-
-Most values have been renamed in the new v2 chart.
+## List of renamed values
 
 | V1 value | V2 value |
 | ----------- | ----------- |
@@ -70,14 +70,15 @@ You may now tune the performance of the driver by setting the following values:
 
 | Name | Description |
 | -------- | ----------- |
-| `controller.readStatusInterval` | The interval between consecutive volume/snapshot checks, raise if you see throttling errors in ReadSnapshot/ReadVolumes calls. |
-| `sidecars.timeout` | The maximum time a sidecar (provisioner, attacher, resizer, snapshotter) will wait for the CSI driver to process a query. Safe to raise if your volumes/snapshots are very large and you encounter timeouts. |
+| `controller.readStatusInterval` | The interval between consecutive volume/snapshot checks; raise it if you see throttling errors in ReadSnapshot/ReadVolumes calls. |
+| `sidecars.timeout` | The maximum time a sidecar (provisioner, attacher, resizer, snapshotter) will wait for the CSI driver to process a query. Raise it if your volumes/snapshots are very large and you encounter timeouts. |
 | `sidecars.kubeAPI.QPS` | The maximum of requests per seconds that a sidecar may make to the Kubernetes API server. |
 | `sidecars.kubeAPI.burst` | The burst above `sidecars.kubeAPI.QPS` allowed for short periods of time. |
-| `sidecars.provisioner.workerThreads` | The number of simultaneous provisioning requests the provisioner sidecar can process. |
-| `sidecars.attacher.workerThreads` | The number of simultaneous attachment requests the attacher sidecar can process. |
-| `sidecars.resizer.workerThreads` | The number of simultaneous resizing requests the resizer sidecar can process. |
+| `sidecars.provisioner.workerThreads` | The number of simultaneous volume provisioning requests the provisioner sidecar can process. |
+| `sidecars.attacher.workerThreads` | The number of simultaneous volume attachment requests the attacher sidecar can process. |
+| `sidecars.resizer.workerThreads` | The number of simultaneous volume resizing requests the resizer sidecar can process. |
 | `sidecars.snapshotter.workerThreads` | The number of simultaneous snapshot requests the snapshotter sidecar can process. |
 
 > Please be aware that there is a limit in the number of API calls you are allowed to make to the Outscale API.
 If you raise `workerThreads` too much, you may decrease the performance of the CSI driver by being throttled by the Outscale API.
+Because throttling limits are applied per call, raising `controller.readStatusInterval` will not increase volume and snapshot creation throughput.