On 2016/10/13 17:19, Yibin Wang wrote:
Thanks for your reply. Comments inline...
On 2016/10/9 23:16, Christoph Hellwig wrote:
- For regisetr, we should stop iteration on failure, and
followed by a
non-stopping unregister operation.
What do you mean with "non-stopping"? Currently once register failed
for a path we then ungerigster all paths, and ignore failures (e.g. due
to a down path or an not already registered path).
BTW, did you mean using dm_pr::fail_early as the indication whether
multipath_iterate_devices() should fail early upon failure? If that's
I can't see it is used anywhere except in dm_pr_register().
Please correct me if I am wrong. Thanks!
That's exactly what I meant - do not stop on failures while doing
dm-devel mailing list