Migration is renaming ethernet connectors

[matheusafernandes]

The migration failed but when we tried to connect again through SSH we saw that the IPv4 was not attached to the eth0. In the "IP a" we see that the eth0 doesn't exist, the ethernet adapter was renamed to ens4. And we have to rebuild the network to get the files.

centos2alma_feedback.zip centos2alma.log

[SandakovMM]

Based on centos2alma logs I could see dnf has been failed on update:

2024-03-27 16:42:46,901 - INFO - Running: ['/usr/bin/dnf', '-y', 'update']. Output:
2024-03-27 16:42:47,142 - INFO - Traceback (most recent call last):
2024-03-27 16:42:47,143 - INFO - File "/usr/bin/dnf", line 57, in <module>
2024-03-27 16:42:47,145 - INFO - from dnf.cli import main
2024-03-27 16:42:47,145 - INFO - File "/usr/lib/python2.7/site-packages/dnf/__init__.py", line 30, in <module>
2024-03-27 16:42:47,146 - INFO - import dnf.base
2024-03-27 16:42:47,146 - INFO - File "/usr/lib/python2.7/site-packages/dnf/base.py", line 29, in <module>
2024-03-27 16:42:47,147 - INFO - import libdnf.transaction
2024-03-27 16:42:47,147 - INFO - File "/usr/lib64/python2.7/site-packages/libdnf/__init__.py", line 3, in <module>
2024-03-27 16:42:47,148 - INFO - from . import conf
2024-03-27 16:42:47,148 - INFO - File "/usr/lib64/python2.7/site-packages/libdnf/conf.py", line 17, in <module>
2024-03-27 16:42:47,149 - INFO - _conf = swig_import_helper()
2024-03-27 16:42:47,149 - INFO - File "/usr/lib64/python2.7/site-packages/libdnf/conf.py", line 16, in swig_import_helper
2024-03-27 16:42:47,149 - INFO - return importlib.import_module('_conf')
2024-03-27 16:42:47,149 - INFO - File "/usr/lib64/python2.7/importlib/__init__.py", line 37, in import_module
2024-03-27 16:42:47,150 - INFO - __import__(name)
2024-03-27 16:42:47,150 - INFO - ImportError: No module named _conf
2024-03-27 16:42:47,168 - ERROR - Command '['/usr/bin/dnf', '-y', 'update']' failed with return code 1

Don't you have any customization in yum configuration?

[matheusafernandes]

Hello @SandakovMM,

This failure on dnf update can happen because as I reported earlier, the VS was operating without IPv4. After all, the migration tool changed the ethernet adapters from eth0 to ens4 pattern. So, the root of the problem is this naming change of the ethernet adapters, which makes the VS cannot connect to the internet to make this update.

1: What can I do to still migrate this Virtual Server? 2: Can you please review why the migration tool is changing the ethernet adapters as a bug?

Thanks.

[SandakovMM]

As far as I understand, leapp does not ensure the stability of NIC interface names. Details could be found on the red hat customer portal. However leapp allows to perform the conversion if there is only one NIC interface. So I expect them to prevent problems in this case.