Douglas Wade Needham wrote:
Quoting Michael DeHaan (mdehaan(a)redhat.com):
> Yes, please let me know details and steps to reproduce this. If you
> have any tracebacks seen in /var/lib/cobbler/cobbler.log or in the Web
> app itself, that would be helpful to see.
>
Reproduction is simple. When clicking on a system instance existing
pre-1.3.1, I get the following traceback...
Looks like the upgrade process needs to make sure the bonding attribute
is applied to loaded but-not-resaved-yet systems.
Generally the cobbler upgrade process calls "cobbler reserialize"
(undocumented) to rewrite the configuration and apply these upgrades, so
folks won't have to manually upgrade things.
Thanks, I'll chase this one down.
Mod_python error: "PythonHandler index"
Traceback (most recent call last):
File "/usr/lib/python2.4/site-packages/mod_python/apache.py", line 299, in
HandlerDispatch
result = object(req)
File "/var/www/cobbler/web/index.py", line 123, in handler
content = func( **form )
File "/usr/lib/python2.4/site-packages/cobbler/webui/CobblerWeb.py", line
478, in system_edit
return self.__render( 'system_edit.tmpl', {
File "/usr/lib/python2.4/site-packages/cobbler/webui/CobblerWeb.py", line
99, in __render
return str(tmpl)
File "/usr/lib/python2.4/site-packages/Cheetah/Template.py", line 994, in
__str__
def __str__(self): return getattr(self, mainMethName)()
File "/usr/lib/python2.4/site-packages/cobbler/webui/master.py", line 222,
in respond
self.body(trans=trans)
File "_usr_share_cobbler_webui_templates_system_edit_tmpl.py", line 397, in
body
KeyError: 'bonding'
Here is the `cobbler system report` for that system:
system : cyteen
profile : fc10-x86_64
comment :
created : Sat Dec 6 12:00:20 2008
image :
kernel options : {}
kernel options post : {}
kickstart : <<inherit>>
ks metadata : {}
mgmt classes : []
modified : Sat Dec 6 12:17:51 2008
netboot enabled? : False
owners : ['cinnion']
server : <<inherit>>
template files : {}
virt cpus : <<inherit>>
virt file size : <<inherit>>
virt path : <<inherit>>
virt ram : <<inherit>>
virt type : <<inherit>>
power type : ipmi
power address :
power user :
power password :
power id :
interface : eth0
bonding :
bonding_master :
bonding_opts :
dhcp tag : cyteen
gateway :
hostname : cyteen
ip address :
is static? : False
mac address : 00:18:f3:53:50:5a
subnet :
virt bridge :
And here is the config file for that system:
---
comment: ''
ctime: 1228582820.01
depth: 2
image: ''
interfaces:
eth0:
dhcp_tag: cyteen
gateway: ''
hostname: cyteen
ip_address: ''
mac_address: '00:18:f3:53:50:5a'
static: False
subnet: ''
virt_bridge: ''
kernel_options: {}
kernel_options_post: {}
kickstart: <<inherit>>
ks_meta: {}
mgmt_classes: []
mtime: 1228583871.93
name: cyteen
netboot_enabled: False
owners:
- cinnion
parent: ''
power_address: ''
power_id: ''
power_pass: ''
power_type: ipmi
power_user: ''
profile: fc10-x86_64
server: <<inherit>>
template_files: {}
virt_bridge: <<inherit>>
virt_cpus: <<inherit>>
virt_file_size: <<inherit>>
virt_path: <<inherit>>
virt_ram: <<inherit>>
virt_type: <<inherit>>
If you want, drop me an email and I will get you copies of my entire
config (/etc/cobbler and /var/lib/cobbler) if that would help. But I
think the key might be that I do not have the 'bonding' field in the
interface, and that we are somehow not getting it initialized.
I think I know what the fix is but I'll let you know. Thanks again!
> Templates resolving on the preservation of old variables like
> $hostname_intf0 (or $hostname_eth0, etc) and the like will need
> tweaking. However, the code should be (and I'll audit this next week)
> set up so that a good value is set for hostname if not already so
> $hostname will continue to work. The new templates cobbler will ship
> with have stuff to set up the networking aspects surrounding these
> automagically, so they would be a good resource to consult.
>
Will do. I figured in part I was just perhaps a little impatient
trying to get a couple of machines at home upgraded to FC10, using the
lastest cobbler to test it, and missed something. Score one for
MythTV. LOL.
You have me confused with the MythTV reference, I hope this mailing list
is not coming in over public access :)
> Thanks!
>
Thanks to you and the others who have been developing cobbler. I have
found it really helpful in the past few months tweeking/reproducing
installations, and hope to convince my supervisor and a coworker at my
work (a major CS and engineering school) that it is worth using for
the data center of our research lab. If all goes well, after the
first of the year, cobbler should be getting a huge workout as we gain
several hundred new hosts.
Hopefully I can speak for everyone -- you're very welcome. If you spot
anything you'd like to see or
encounter any other bugs, please let us know.
--Michael
- Doug