Bug 462083
| Summary: | Cannot retrieve kickstart files over the newtork | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Orion Poplawski <orion> | ||||
| Component: | anaconda | Assignee: | David Cantrell <dcantrell> | ||||
| Status: | CLOSED RAWHIDE | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | medium | ||||||
| Version: | rawhide | CC: | hdegoede, jlaska, jon.wallace, oliva, wwoods, ydesgagn | ||||
| Target Milestone: | --- | Keywords: | Reopened | ||||
| Target Release: | --- | ||||||
| Hardware: | All | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2008-10-03 22:44:55 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: | |||||||
| Bug Depends On: | |||||||
| Bug Blocks: | 438943 | ||||||
| Attachments: |
|
||||||
|
Description
Orion Poplawski
2008-09-12 14:55:08 UTC
Can't reproduce this with the latest nightly. I think the missing info is that this is a pxeboot/kickstart install: kernel fedora/development/i386/vmlinuz append root=/dev/ram initrd=fedora/development/i386/initrd.img ks=nfs:saga:/export/data1/ks/rawhide-desk.cfg Same for http. HTTP installs work fine here. It's gotta be PXE-specific. Still seeing with 11.4.1.39. Still the same with 11.4.1.40. Let me know if there is anything else I can do to help debug. So I guess NM fails to come up and the system never fetches its kickstart? Are there any useful messages on VT2/VT3? Have you tried adding network configuration directives to the commandline, like "ip=dhcp ipv6=auto"? How did the network get configured when doing PXE-boot installs in the past? (In reply to comment #6) > So I guess NM fails to come up and the system never fetches its kickstart? Correct. > Are there any useful messages on VT2/VT3? Not really other than those mentioned above. > Have you tried adding network configuration directives to the commandline, like > "ip=dhcp ipv6=auto"? Then I get: INFO: getting kickstart file ERROR: nm_ip2str returned NULL And I get a text box prompt to enter a new kickstart location > How did the network get configured when doing PXE-boot installs in the past? dhcp - but probably directly controlled from anaconda, not NM. Okay, recompiled loader that will spit out debug messages and get: DEBUG : get_connection (1804): org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.NetworkManager was not provided by and .service files Looks like some issues with how NM is started. With getting the kickstart file over the network, we need to bring the network up very early.
In a non-network-kickstart file case NM is started up by iface_start_NetworkManager() in isys/iface.c, but control files are needed:
/*
* Start NetworkManager -- requires that you have already written out the
* control files in /etc/sysconfig for the interface.
*/
int iface_start_NetworkManager(DBusConnection *connection, char **error) {
but nothing is writing ifcfg scripts at the retrieve kickstart point.
Created attachment 317821 [details]
Patch to bring up NM early
I was wrong, you are writing out the ifcfg file, but not starting NM. The attached patch works for me.
*** Bug 463409 has been marked as a duplicate of this bug. *** I ran into this on F10-Beta. *** Bug 465314 has been marked as a duplicate of this bug. *** I just committed a patch fixing this, so the next anaconda build (11.4.1.43-1) should have this fixed. *** Bug 465573 has been marked as a duplicate of this bug. *** *** Bug 468456 has been marked as a duplicate of this bug. *** |