This bug has been migrated to another issue tracking site. It has been closed here and may no longer be being monitored.

If you would like to get updates for this issue, or to participate in it, you may do so at Red Hat Issue Tracker .
Bug 2067171 - Migration is failing during disk conversion on RPM guests using non-BDB
Summary: Migration is failing during disk conversion on RPM guests using non-BDB
Keywords:
Status: CLOSED MIGRATED
Alias: None
Product: Migration Toolkit for Virtualization
Classification: Red Hat
Component: General
Version: 2.3.0
Hardware: Unspecified
OS: Unspecified
high
urgent
Target Milestone: ---
: ---
Assignee: Arik
QA Contact: Ilanit Stein
Richard Hoch
URL:
Whiteboard:
Depends On:
Blocks: 2067177
TreeView+ depends on / blocked
 
Reported: 2022-03-23 13:09 UTC by Igor Braginsky
Modified: 2023-01-02 13:16 UTC (History)
5 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2023-01-02 13:13:54 UTC
Target Upstream Version:
Embargoed:

Attachments (Terms of Use)
must-gather logs of failed plan (33.97 KB, application/gzip)
2022-03-23 13:09 UTC, Igor Braginsky 		no flags Details
View All


Links
System ID Private Priority Status Summary Last Updated
Github konveyor forklift-virt-v2v pull 5 0 None Merged Replace centos8 EOL base image in favor of stream8 2022-07-07 12:54:27 UTC
Red Hat Issue Tracker   MTV-444 0 None None None 2023-01-02 13:16:28 UTC

Description Igor Braginsky 2022-03-23 13:09:38 UTC 
Created attachment 1867804 [details]
must-gather logs of failed plan

Description of problem:Migration is failing on disk conversion step if source VM is UEFI


Version-Release number of selected component (if applicable): 2.3.0-38


How reproducible: 100%


Steps to Reproduce:
1. Migrate UEFI VM from VMWare
2.
3.

Actual results:
Migration is failing on disk conversion

Expected results:


Additional info:
Gathered logs with must-gather and attached here
Comment 1 Matthew Arnold 2022-04-14 17:14:01 UTC 
It looks like this is not due to UEFI itself. I was able to migrate a few RHEL7 VMs, with virt-v2v successfully detecting UEFI firmware. The error from the log is:


libguestfs: trace: v2v: filesize "/var/lib/rpm/Name"
_package_format (0x1e5) took 0.01 secs
guestfsd: <= filesize (0xda) request length 64 bytes
guestfsd: error: /var/lib/rpm/Name: No such file or directory
guelibguestfs: trace: v2v: filesize = -1 (error)
libguestfs: trace: v2v: inspect_list_applications2 = NULL (error)
virt-v2v: error: libguestfs error: filesize: /var/lib/rpm/Name: No such 
file or directory
rm -rf '/var/tmp/null.TlX9xM'


I was assuming the failing VM was RHEL7 (since the name was uefi-mtv-rhel7), but I downloaded the disk image and it's a recent version of Fedora. For this to work, we need the forklift-virt-v2v image to pick up this fix: https://github.com/libguestfs/libguestfs/commit/c9ee831affed55abe0f928134cbbd2ed83b2f510

I will see if I can get that image updated.
Comment 2 Matthew Arnold 2022-04-20 20:23:39 UTC 
Unfortunately it turns out the fix is more than a simple container image update, it would need a Fedora 36 image and a rework of how forklift uses virt-v2v in place. That might be a big enough change to justify pushing this bug off 2.3.1, but the cost is that we will not be able to migrate RHEL9 VMs in 2.3.1. I have migrated RHEL/CentOS 7 up to 8.3, but I expect RHEL9 VMs to fail with this same error (whether or not they use UEFI). I'm not sure how high of a priority RHEL9 support is for MTV at the moment.
Comment 3 Ilanit Stein 2022-06-08 14:57:30 UTC 
Targeting to MTV-2.3.3 to assess it then.
Comment 4 Jeff Ortel 2022-07-06 16:53:06 UTC 
This has nothing to do with UEFI and does not affect RHEL8.
The failure in the bug is from a change in the RPM database format that our version of virt-v2v doesn't handle right.
Updating target release to FUTURE.
Comment 5 Matthew Arnold 2022-07-07 12:54:27 UTC 
I didn't realize the libguestfs fix had made it into CentOS Stream 8, and I didn't notice Franco had updated forklift-virt-v2v to centos:stream8 on the same day I made my last comment. So this should actually already be fixed for 2.3.2.
Comment 6 Arik 2023-01-02 13:16:29 UTC 
The fix should already be available, I'm migrating this bz to Jira just to make sure it does
Note You need to log in before you can comment on or make changes to this bug.