Skip to contentSkip to conversation

Fix 'bz 599898 double convert of mirror ignores log option'

PATCH New Jonathan Brassow 1 participant 1 message Jun 17, 2010

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Fix 'bz 599898 double convert of mirror ignores log option'

As well as opinions on the patch, I'm hoping for comments on whether it is the right approach. There are really 2 options:
1) disallow log conversions while mirrors are stacked
2) fix the issue, because the problem is not /that/ ambiguous just a little tricky perhaps.
I've taken the second approach.

brassow

Mirrors can be layered - as in the case of an converting 2-way to 3-way mirror. When conversion operations are performed on these types of mirrors, log options can be confused/ignored.

In the case of a converting 3-way mirror, we have a top-level 2-way corelog mirror whose legs are 1) a 2-way disk-log mirror and 2) a linear device. If we wish to convert this 3-way mirror to a 2-way mirror, the linear device is removed and the extra top layer is eliminated. If we also wished to convert the disk log to a core log in the same step, ambiguity creeps in. It is somewhat obvious what the user wants - a 2-way mirror with a corelog. However, looking at the top level mirror before compression, it seems that the mirror already has a core log. This is why the operation seemed to fail.

This patch simply re-evaluates what mirrored_seg points to after a compression and then considers the log argument.

This is a fix for bug 599898.

lib/metadata/mirror.c +3 -0
Index: LVM2/lib/metadata/mirror.c
===================================================================
@@ -851,6 +851,9 @@ static int _remove_mirror_images(struct 
 			log_error("Failed to add mirror images");
 			return 0;
 		}
+		mirrored_seg = first_seg(lv);
+		if (remove_log && !detached_log_lv)
+			detached_log_lv = detach_mirror_log(mirrored_seg);
 	} else if (new_area_count == 0) {
 		log_very_verbose("All mimages of %s are gone", lv->name);