-
Bug
-
Resolution: Won't Do
-
Major
-
None
-
rhel-8.8.0
-
None
-
Important
-
rhel-bootloader
-
ssg_core_services
-
8
-
False
-
False
-
-
None
-
None
-
None
-
None
-
If docs needed, set a value
-
-
All
-
None
-
57,005
Description of problem:
This is a continuation of dracut BZ #2212804.
When customers try to install systems over the network and specify a malformed ip= kernel command line parameter, the dracut emergency prompt is eventually reached after a very long time (~1 hour), instead of somehow expected 4/5 minutes.
During all that time, we can only see the following message regularly (every ~10 seconds):
-------- 8< ---------------- 8< ---------------- 8< ---------------- 8< --------
[ 642.665130] dracut-initqueue[5561]: Warning: Downloading 'http://192.168.122.1/rhel88/images/install.img' failed!
[ 642.670886] dracut-initqueue[5576]: % Total % Received % Xferd Average Speed Time Time Time Current
[ 642.670943] dracut-initqueue[5576]: Dload Upload Total Spent Left Speed
[ 645.729126] dracut-initqueue[5576]: 0 0 0 0 0 0 0 0 -::- -::- -::- 0 0 0 0 0 0 0 0 0 -::- 0:00:01 -::- 0 0 0 0 0 0 0 0 0 -::- 0:00:02 -::- 0curl: (7) Failed to connect to 192.168.122.1 port 80: No route to host
[ 645.738017] dracut-initqueue[5570]: Warning: Downloading 'http://192.168.122.1/rhel88/LiveOS/squashfs.img' failed!
[ 645.738805] dracut-initqueue[940]: Warning: anaconda: failed to fetch stage2 from http://192.168.122.1/rhel88
[ 648.823442] dracut-initqueue[940]: Warning: can't find installer main image path in .treeinfo
-------- 8< ---------------- 8< ---------------- 8< ---------------- 8< --------
The issue also happens if the ip= is properly specified but the network is not functional, e.g. because the network requires a VLAN specification but it was not specified on kernel command line.
We need something to drop early, e.g. some absolute timeout.
Version-Release number of selected component (if applicable):
dracut-049
anaconda-dracut-33.16.8.9
How reproducible:
Always
Steps to Reproduce:
1. Setup a QEMU/KVM with isolated network (no DHCP)
2. Boot by specifying kernel and initrd paths and "ip=someip::... inst.repo=xxx" as kernel args
(repo pointing to unreachable network)
Actual results:
After 900 seconds, still at loop 90 (over 240)
Expected results:
After 180 seconds (the rd.retry default), getting the dracut prompt
Additional info:
This is critical for a lot of customers, because tuning network install snippets is not that easy.
- external trackers