| Summary: | Network HTTP Installer fails consistently trying to find media.cfg in stage 2 | ||
|---|---|---|---|
| Product: | Mageia | Reporter: | Frank Griffin <ftg> |
| Component: | Installer | Assignee: | Mageia tools maintainers <mageiatools> |
| Status: | NEW --- | QA Contact: | |
| Severity: | major | ||
| Priority: | Normal | CC: | isobuild, mageia, marja11 |
| Version: | Cauldron | ||
| Target Milestone: | --- | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Source RPM: | CVE: | ||
| Status comment: | |||
| Attachments: | report.bug.xz from the failure | ||
|
Description
Frank Griffin
2023-02-05 22:56:38 CET
Created attachment 13677 [details]
report.bug.xz from the failure
(In reply to Frank Griffin from comment #0) > This relates to bug#24362. As the Summary says. However, the point of a > new bug is that it fails on the same file. This is not the normal flaky > network error that a wget rc 4 indicates. The network connection is an > intranet connection which has no problems otherwise. I've tried this twice, > and both times it's media.cfg that fails. That tells me that either stage 2 > has lost the network or its DNS by the time the HTTP request for media.cfg > is issued, or else there is something wrong with the wget parameters being > used. Thanks for the report. From the logs * change_phys_medium http://ftgfiles1/mnt/cauldron/x86_64/media for file media_info * getFile media_info/media.cfg on http://ftgfiles1/mnt/cauldron/x86_64/media * retrieving media.cfg * '/usr/bin/wget' '--force-clobber' '-t' '3' '--retr-symlinks' '--timeout=60' '-P' '/mnt/var/cache/urpmi' 'http://ftgfiles1/mnt/cauldron/x86_64/media/media_info/media.cfg' | * error: wget failed: exited with 4 Assigning to the Mageia Tools maintainers, CC'ing the ISO builders CC:
(none) =>
isobuild, marja11 Can't reproduce. Tested using the Mageia-Cauldron-netinstall-nonfree-x86_64.iso dated Feb 6 17:35, over both wired and wireless connections to both a local and remote mirror.
In your report.bug.xz (attachment 13677 [details]) the syslog section shows many repeats of
<6>[ 97.540627] pcieport 0000:00:1c.3: AER: Multiple Corrected error received: 0000:02:00.0
<4>[ 97.540649] rtw_8821ce 0000:02:00.0: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
<4>[ 97.540652] rtw_8821ce 0000:02:00.0: device [10ec:c821] error status/mask=00000001/0000e000
<4>[ 97.540655] rtw_8821ce 0000:02:00.0: [ 0] RxErr
which suggests all is not well with your network connection.CC:
(none) =>
mageia It's possible, but I don't think so. The cauldron mirror and the installing host are both on the same intranet connected by the same switch and the host using the same switch through a wireless router. This arrangement has several copies of the mirror which are updated multiple times a day from the primary copy through the same network path with no errors via rsync. There are other old bugs about this, but it would be really great if we could get other network installs like NFS working to get HTTP out of the loop. I retried this using wired ethernet via USB, and it ran fine. But nothing in my house besides our installer has any problem with the wireless, and this machine has no problem with the same wireless running an installed cauldron system. So far now, the installer has had problems with both curl and wget. I'm thinking it has to be the way we're using the utilities. |