Fixing Spacewalk Issues

OSA Dispatcher

Symptoms:

1. Log /var/log/rhn/osa-dispatcher.log running full

2. Log /var/log/rhn/osa-dispatcher.log shows these lines:

self.verify_peer(ssl)\n  File "/usr/lib/python2.7/site-packages/osad/jabber_lib.py", line 741, in verify_peer\n    (self._host, common_name))\nSSLVerifyError: Mismatch: peer name:

Resolution possibilities:

1. Re-run the last upgrade.

2. Update /etc/rhn/rhn.conf to include the full hostname  on
server.jabber_server
osa-dispatcher.jabber_server

If you change cobbler.host, check via the web interface, whether the tab "Provisioning" still works or throws an error.

Symptoms:

/var/log/rhn/osa-dispatcher.log: jabberd/c2s[24205]: SASL callback for non-existing host: spacewalk.example.com

/var/log/rhn/osa-dispatcher.log jabberd/c2s[11498]: sqlite (authreg): no such table: authreg

Resolution Possibilities

Update /etc/jabberd/c2s.xml and replace hostname with the FQDN

# sqlite3 /var/lib/jabberd/db/sqlite.db < /usr/share/jabberd/db-setup.sqlite
# chown jabber:jabber /var/lib/jabberd/db/sqlite.db

 

Type "hostname". It should be the FQDN. Make sure it is also in /etc/sysconfig/network as

HOSTNAME=spacewalk.example.com

spacewalk-setup-jabberd
spacewalk-control restart

Make sure IPv6 is enabled

 

https://access.redhat.com/solutions/327903

Internal Error on listing Scheduled Actions

Symptoms:

You get an internal error when clicking on the details of "Scheduled Actions" --> "Failed Actions" or "Completed Actions"

Resolution Possibilities

This is a reported bug in https://bugzilla.redhat.com/show_bug.cgi?id=1498260

Try restarting the spacewalk server via the web interface: /rhn/admin/config/Restart.do
If that did not work, try the next steps or check the bugzilla for other options

systemctl stop tomcat
systemctl stop taskomatic
systemctl stop rhn-search
systemctl start rhn-search
systemctl start taskomatic
systemctl start tomcat
spacewalk-service restart

If this did not work, a. Add line to rhn.conf file and try the above commands again.

hibernate.enable_lazy_load_no_trans=true

This might have to be done after each server restart.

Kickstart fails

Symptoms:

On the client during the Kickstart startup: "Installation Source" - "Error setting up software source".
On the server: GET /rhn/common/DownloadFile.do fails with a null pointer exception.

Resolution Possibilities

https://bugzilla.redhat.com/show_bug.cgi?id=1323548
https://bugzilla.redhat.com/show_bug.cgi?id=1713283

Set up your channel hierarchy properly. Don't use the root node as channel but as a group placeholder.

Also untick the box "Always use the newest Tree for this base channel. "Newest" is determined by the date it was last modified."

Edit .treeinfo in the kickstart folder and  remove the variants references to other repositories (i.e. Appstream for CentOS 8)

 

Clients are not checking in

Symptoms:

The spacewalk system overview shows question marks for your clients and never update.

Resolution Possibilities

https://bugzilla.redhat.com/show_bug.cgi?id=1688278

I am assuming that you already checked that your client is switched on :). Access the client and make sure that rhnsd.timer is running and rhnsd.service is not:
systemctl enable rhnsd.timer
systemctl start rhnsd.timer

If rhnsd.timer does not exist, install the package rhnsd.

 

Creating Virtual Machines fail

Symptoms:

On the client oyu can see in then rhn_check logs that virt-install fails with an unknown location.

Resolution Possibilities

The kickstart distribution does not have a .treeinfo (this file is not downloaded by the default spacewalk sync. See https://github.com/spacewalkproject/spacewalk/pull/702). Download the .treeinfo file manually to your kickstart distribution folder or integrate the pull request. Note the kickstart issue above!

Client cannot find packages

Symptoms:

Running updates on the client returns messages in rhn_check and yum/dnf that packages cannot be downloaded.
When pushing packages from the server, the packages keep showing up as updates.

Resolution Possibilities

On client: Clear the cache by running
yum clean all # or dnf clean all
or as last resort, delete these folders:
rm -Rf /var/cache/yum/*
rm -Rf /var/cache/dnf/*

On Server: Make sure that taskomatic is running. Maybe restart the service if it got stuck. Taskomatic is responsible to create the repomd.xml file for all new imported packages. If this did not run (properly) after importing the packages, clients cannot find it.