asomers commented on this pull request.
> c = vd->vdev_children;
mm = kmem_zalloc(offsetof(mirror_map_t, mm_child[c]), KM_SLEEP);
mm->mm_children = c;
- mm->mm_replacing = (vd->vdev_ops == &vdev_replacing_ops ||
+ /*
+ * If we are resilvering, then we should handle scrub reads
+ * differently; we shouldn't issue them to the resilvering
+ * device because it might not have those blocks.
+ *
+ * We are resilvering iff:
+ * 1) We are a replacing vdev (ie our name is "replacing-1" or
+ * "spare-1" or something like that), and
+ * 2) The pool is currently being resilvered.
+ *
+ * We cannot simply check vd->vdev_resilvering, because that
I guess a lot changes in 3.5 years. I just did a little experiment, and the
replacement `vdev_resilver_txg` is also not set in this case. I'll update the
comment.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/422#discussion_r135370900
------------------------------------------
openzfs-developer
Archives:
https://openzfs.topicbox.com/groups/developer/discussions/Te100c9e22da2b1d8-M4c8688f2c95371d92fa84a83
Powered by Topicbox: https://topicbox.com
