Allow a custom driver name to be specified on the commandline #179
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This was referenced Sep 13, 2019
adamwg
commented
Sep 13, 2019
| url = flag.String("url", "https://api.digitalocean.com/", "DigitalOcean API URL") | ||
| doTag = flag.String("do-tag", "", "Tag DigitalOcean volumes on Create/Attach") | ||
| version = flag.Bool("version", false, "Print the version and exit.") | ||
| endpoint = flag.String("endpoint", "unix:///var/lib/kubelet/plugins/"+driver.DefaultDriverName+"/csi.sock", "CSI endpoint") |
There was a problem hiding this comment.
Note: We're using the default driver name here, not the commandline-provided one. The endpoint socket path just needs to match what's in the deployment manifest, not the driver name.
jcodybaker
approved these changes
Sep 16, 2019
driver/node.go
Outdated
| @@ -95,7 +90,7 @@ func (d *Driver) NodeStageVolume(ctx context.Context, req *csi.NodeStageVolumeRe | |||
| "method": "node_stage_volume", | |||
| }) | |||
|
|
|||
| _, ok := req.VolumeContext[annNoFormatVolume] | |||
| _, ok := req.VolumeContext[d.annNoFormatVolume] | |||
There was a problem hiding this comment.
I wonder if should allow both old and new annotations here. That might make user-facing documentation easier and reduce the pain caused by confusing the two.
There was a problem hiding this comment.
Good idea. I've implemented this in the latest push.
I hard-coded the new and legacy driver names for this rather than using the flag - seems sensible given that we're supporting both, but let me know if you have a nicer way to do it.
The name of our CSI driver changed between the 0.2 and 0.4 releases, from `com.digitalocean.csi.dobs` to `dobs.csi.digitalocean.com`. This means that when upgrading a cluster from Kubernetes 1.11 (which uses CSI plugin 0.2) to Kubernetes 1.12 (which uses CSI plugin 0.4) volumes created before the upgrade have the old driver name in various annotations and spec fields. This prevents the new driver from managing the volume, leading to broken workloads. To handle the upgrade case, allow a custom driver name to be specified on the commandline. This way, a cluster that was once running 1.11 and thus uses the old name can continue using the old name. By default the new name will be used.
adamwg
force-pushed
the
awg/custom-driver-name-master
branch
from
63f14a8 to
221481b
Compare
snormore
approved these changes
Sep 17, 2019
| // This annotation is added to a PV to indicate that the volume should be | ||
| // not formatted. Useful for cases if the user wants to reuse an existing | ||
| // volume. | ||
| annNoFormatVolume = DriverName + "/noformat" | ||
| // volume. We support using either the legacy driver name |
adamwg
added a commit
that referenced
this pull request
Sep 17, 2019
In #179 we changed the `noformat` annotation to `no-format`, which was not intentional. Change it back so it works as before.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The name of our CSI driver changed between the 0.2 and 0.4 releases, from
com.digitalocean.csi.dobstodobs.csi.digitalocean.com. This means that when upgrading a cluster from Kubernetes 1.11 (which uses CSI plugin 0.2) to Kubernetes 1.12 (which uses CSI plugin 0.4) volumes created before the upgrade have the old driver name in various annotations and spec fields. This prevents the new driver from managing the volume, leading to broken workloads.To handle the upgrade case, allow a custom driver name to be specified on the commandline. This way, a cluster that was once running 1.11 and thus uses the old name can continue using the old name. By default the new name will be used.