Use the filesystem from the object cache when unmounting / mounting filesystems. #1978
Conversation
filesystems. Signed-off-by: johnsonw <wjohnson@whamcloud.com>
| "host": self.host, | ||
| "filesystems": [ | ||
| ( | ||
| ObjectCache.get_one(ManagedFilesystem, lambda mf, mtd=m: mf.name == mtd.filesystem).mount_path(), |
There was a problem hiding this comment.
What's the difference comparing to the above two lines (search= and unmounted=)?
There was a problem hiding this comment.
The unmounted is a list of unmount LustreClientMount objects. We loop through each of these and match on the filesystem whose name in the umounted client mount objects matches the filesystem name. That will then give a list of filesystems that correspond to the mounts. For example, say we have:
Filesystems:
fs1, fs2, and fs3.
Mounts:
mnt1, mnt2
mnt1 is associated with fs1
mnt2 is associated with fs2
We unmount both of these instances but their object still exists. We then later want to mount them. The expression would return a list with:
[fs1 mount path, fs2 mount path]
There was a problem hiding this comment.
The problem i'm working through at the moment is that the mount path is erased when a client is unmounted.
There was a problem hiding this comment.
That can (should) be in a separate patch.
…1978) filesystems. Signed-off-by: johnsonw <wjohnson@whamcloud.com>
Use the filesystem from the object cache when unmounting / mounting filesystems.
Signed-off-by: johnsonw wjohnson@whamcloud.com
This change is