Bug 506183
Summary: | unable to recreate failed log device after successful convert to core log | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | Red Hat Enterprise Linux 5 | Reporter: | Corey Marthaler <cmarthal> | ||||||||
Component: | lvm2 | Assignee: | Petr Rockai <prockai> | ||||||||
Status: | CLOSED DEFERRED | QA Contact: | Cluster QE <mspqa-list> | ||||||||
Severity: | medium | Docs Contact: | |||||||||
Priority: | medium | ||||||||||
Version: | 5.4 | CC: | agk, dwysocha, edamato, heinzm, jbrassow, mbroz, prockai | ||||||||
Target Milestone: | rc | ||||||||||
Target Release: | --- | ||||||||||
Hardware: | All | ||||||||||
OS: | Linux | ||||||||||
Whiteboard: | |||||||||||
Fixed In Version: | Doc Type: | Bug Fix | |||||||||
Doc Text: | Story Points: | --- | |||||||||
Clone Of: | Environment: | ||||||||||
Last Closed: | 2009-11-10 17:06:35 UTC | Type: | --- | ||||||||
Regression: | --- | Mount Type: | --- | ||||||||
Documentation: | --- | CRM: | |||||||||
Verified Versions: | Category: | --- | |||||||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||||
Cloudforms Team: | --- | Target Upstream Version: | |||||||||
Embargoed: | |||||||||||
Attachments: |
|
Description
Corey Marthaler
2009-06-15 22:18:51 UTC
Created attachment 348021 [details]
archive 1
Created attachment 348022 [details]
archive 2
Created attachment 348023 [details]
archive 3
The pvcreate says to use -ff to force overwriting a device with existing LVM metadata on it. If you fail a device, the device is of course not wiped by LVM (it is gone) nor when it returns (doing this would likely lead to data loss in at least some cases). When it comes back, it still has the old metadata on it, so LVM will, by default, refuse to pvcreate. In this case, however, the new version of the mirror repair patch should also run vgreduce --removemissing (without --force), which would lead to a clean volume group (since there is nothing in the VG besides the partial mirror). I guess we want to track this bug for the new mirror repair code -- we need to double-check that this is really fixed. It seems like the metadata on the devices that stay up are changed to reflect that the failed device is gone, whether or not the metadata on the failed device is ever changed. This test had never failed like that before and hasn't since. I've run this now quite a few times with the latest rpms and haven't seen any issues. This appears to be some kind of fluke failure/bug. lvm2-2.02.46-10.el5 BUILT: Fri Sep 18 09:38:06 CDT 2009 lvm2-cluster-2.02.46-10.el5 BUILT: Fri Sep 18 09:39:48 CDT 2009 Ok. I'l close this as deferred, due to complete lack of reproducibility. If you ever run into that problem again, please let me know. |