[Spacewalk-list] Spacewalk registration/updating issues

[Michael Mraka]

Philip Owens:
> Recently I took over administration of a Spacewalk server that someone else
> built that no longer works for us. It works 95% of the time. However,
> sometimes, after initial registration and successful updating, when one
> goes to update down the road--whether it is security only or a full
> update--I will get an error with rhn_check:
> 
> *[root at XXXXXXXX log]# rhn_checkrhn-plugin: Error communicating with server.
> The message was:*
> *Internal Server Error*
> 
> AND:
> 
> *[root at nsh-olo-dnsmasq-p01 log]# rhn-profile-sync -vvupdateLoginInfo()
> login infoD: login(forceUpdate=True) invokedD: rpcServer: Calling XMLRPC
> registration.upgrade_versionA protocol error occurred: Internal Server
> Error , attempt #1,There was a RhnServerException while testing login:*
> *---> Blank Line <---*
> 
> So when I have seen this, I usually check the apache error log on the
> spacewalk server and I will see this error:
> 
> 
> *[Mon Aug 12 14:03:34.152735 2019] [:error] [pid 3433] TypeError: ('cannot
> marshal None unless allow_none is enabled'*

Hello Philip,

I've seen similar issue in the past when clients had much older packages
than server. E.g. Spacewalk 2.2 clients vs 2.8 server.
I'd try to update client packages.

> Thus, I have to delete the system in Spacewalk and then re-register it with
> the force flag, and then I can resume updates. My question is, why is this
> happening? I might believe this could happen over a long period. However,
> this is happening some times in less than a week. Any help would be
> appreciated. The systems are running OL 7.4, 7.5, or 7.6 when they get
> registered to Spacewalk.

Regards,

--
Michael Mráka
System Management Engineering, Red Hat