[Spacewalk-list] Problem installing Spacewalk on CentOS7 (Brand new install)

Bruce Wainer bruce at brucewainer.com
Mon Dec 11 01:17:17 UTC 2017

After leaving this be a few months I returned to the issue. I found the
cause to be that my hostname had capitalized letters in it. Renaming the
host with all lowercase and running the spacewalk-hostname-rename script
resolved the issue. Apparently, the Satellite 5.8 Installation Guide (
Section 2.4.8) has a note that saying not to do this as it would cause
problems between jabberd and OSAD (exactly what I was seeing) but this
warning is NOT present in the up-to-date install guide for Spacewalk (
https://github.com/spacewalkproject/spacewalk/wiki/HowToInstall ). I think
that needs to be fixed.
Here is the bug report that tipped me off to this issue:

I've never used github aside from downloading releases and contributing to
bug reports, so I don't know how to edit the wiki myself (looks like I have
to make a clone/fork of it, and then a pull request with my changed
version?) If someone wouldn't mind adding a note about using all-lower-case
hostnames, maybe copied directly from the Satellite Install Guide, I think
that would help future users. Nothing in the outputs of jabberd-c2s or osad
help point to this being an issue, you just get errors like "no host
available for requested domain", "SASL callback for non-existing host", and
"bind.c:38 not auth'd, offering auth and register".

Hope this helps someone (maybe a future person searching for a solution to
this issue),

Bruce Wainer

On Wed, Jun 28, 2017 at 1:56 AM, Bruce Wainer <bruce at brucewainer.com> wrote:

> Robert,
> I beg to differ. I have looked that up, and that string appears in every
> single error that osa-dispatcher throws. It appears to be an incorrectly
> written exception handling code, that causes its own exception. It always
> appears right after "('Error caught:',)" which was clearly intended to
> print out a particular error message, but instead fails and causes another
> exception. See:
> https://www.redhat.com/archives/spacewalk-list/2016-June/msg00021.html
> https://www.redhat.com/archives/spacewalk-list/2016-
> September/msg00003.html
> https://bugzilla.redhat.com/show_bug.cgi?id=1336426 (this isn't a bug for
> this issue, just one that has a log that shows this activity)
> http://spacewalk-list.redhat.narkive.com/k0ly5UDb/error-
> connecting-to-jabber
> These are all cases where different osa-dispatcher errors all are followed
> by
> ('Error caught:',)
> ERROR: unhandled exception occurred: (unicode argument expected, got 'str')
> So I have to conclude that this particular message is not the cause of my
> problem. And even if it were, I don't see any way I could have caused or
> fix this error. My hostname doesn't have unicode in it, and expecting
> unicode in a hostname would be very silly (allowing unicode in a hostname
> might be ok, but not requiring it). The fact that there is an unhandled
> exception at all means that there is a bug in the code, but I'm 99% sure
> that is in the Error-handling code and not something that is functional to
> the program. If you really think this is the problem that is affecting me,
> please provide some help as to fixing it, as Google brings up nothing.
> Bruce Wainer
> On Wed, Jun 28, 2017 at 1:22 AM, Robert Paschedag <robert.paschedag at web.de
> > wrote:
>> Jun 27 16:59:05 NETMAN osa-dispatcher: ERROR: unhandled exception
>> occurred: (unicode argument expected, got 'str').
>> That's your reason.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/spacewalk-list/attachments/20171210/00ffb2c0/attachment.htm>

More information about the Spacewalk-list mailing list