Windows: Clarify r/o root filesystem#819
Conversation
Signed-off-by: John Howard <jhoward@microsoft.com>
| On Linux, for example, with a bundle at `/to/bundle` and a root filesystem at `/to/bundle/rootfs`, the `path` value can be either `/to/bundle/rootfs` or `rootfs`. | ||
| A directory MUST exist at the path declared by the field. | ||
| * **`readonly`** (bool, OPTIONAL) If true then the root filesystem MUST be read-only inside the container, defaults to false. | ||
| * **`readonly`** (bool, OPTIONAL) If true then the root filesystem MUST be read-only inside the container, defaults to false. On Windows, this field must be omitted or false. |
There was a problem hiding this comment.
Previous discussion in #23 and #476.
An alternative to this would be to label the property as Linux/Solaris-specific, so it wouldn't be specified for Windows and Windows runtimes could silently ignore it (using this). That gets us to the runtime behavior proposed in #476, but avoids confusing config authors (like the spec wording in #476) because Windows authors would know that the property did not apply to their platform, and non-Windows authors would know that the property did apply (and runtimes would have to implement it) on their platforms.
|
LGTM Forcing this to be a known-supported value like this makes sense IMO -- that allows for it to be supported in a future update if the platform adds support. This also ends up mirroring other places in the spec where features requested in the configuration that are unsupported by the runtime or the underlying kernel MUST result in an error (which is essentially what this is codifying for Windows). |
To clarify my alternative proposal, what do you think a Linux runtime should do if a Linux config sets the Windows-only |
|
As a user, anything other than an error would be very surprising behavior.
|
|
LGTM |
Signed-off-by: John Howard jhoward@microsoft.com
Clarification that Windows does not support a read-only root filesystem.