[Spacewalk-list] no cache rebuild after cleanup from /var/cache/rhn

[Thomas von Steiger]

On 09.04.2009, at 16:05, Pradeep Kilambi wrote:

> Cliff wrote:
>> Thomas von Steiger wrote:
>>> Hi,
>>>
>>> I have stopped spacewalk and removed everthing from /var/cache/rhn.
>>> After restart from spacewalk 0.5 there is no rebuild of the cache.
>>>
>>> From the cent5 clients i get by yum check-update:
>>>
>>> Error: Cannot retrieve repository metadata (repomd.xml)
>>>
>>> The only errors i can found in the log's are from osad/jabberd:
>>>
>>> 2009/04/08 23:37:52 +02:00 17861 0.0.0.0: osad/jabber_lib.__init__
>>> 2009/04/08 23:37:52 +02:00 17861 0.0.0.0: osad/ 
>>> jabber_lib.print_message('socket error',)
>>> 2009/04/08 23:37:52 +02:00 17861 0.0.0.0: osad/ 
>>> jabber_lib.print_message('Could not connect to jabber server',  
>>> 'spacewalk2.dom.local')
>>> 2009/04/08 23:37:52 +02:00 17861 0.0.0.0: osad/ 
>>> jabber_lib.setup_connection('Could not connect to any jabber  
>>> server',)
>>> 2009/04/08 23:37:52 +02:00 17861 0.0.0.0: osad/ 
>>> jabber_lib.main('Unable to connect to jabber servers, sleeping 10  
>>> seconds',)
>>>
>>> [root at spacewalk2 log]# grep cache /etc/rhn/rhn.conf
>>> repomd_cache_mount_point = /var/cache
>>>
>>> What can i do?
>>>
>>> Thomas
>>>
>>>
>
> Just restarting spacewalk will not rebuild the cache. The cache  
> regenration should have started when you're centos client checked in.
>
> check your /var/cache/rhn/repodata and the cache should be generated  
> in a bit. The cache generation for repodata is triggered only when
> there is a change to the channel state(through means like rhnpush,  
> satsync, api or ui) or when a client checks in and cache is missing.
>
> Check back in a bit and cache should be there,
>
> ~ Prad

I have tested this again and found, on all the clients i get this  
error message since update to 0.5.
In the meantime i have reinstalled all the rhn client rpm's from new  
and registerd success for new on the spacewalk.
rhel-instnum is installed on client and server:

[root at cent5n1 rhn]# yum check-update
Loading "rhnplugin" plugin
Loading "fastestmirror" plugin
Loading mirror speeds from cached hostfile
  * centos-5-base-x86_64: spacewalk2.dom.local
  * centos-5-updates-x86_64: spacewalk2.dom.local
  * centos-5-centosplus-x86_64: spacewalk2.dom.local
  * centos-5-extras-x86_64: spacewalk2.dom.local
  * spacewalk-client-tools: stahnma.fedorapeople.org
Error: Cannot retrieve repository metadata (repomd.xml) for  
repository: centos-5-base-x86_64. Please verify its path and try again

On the spacewalk self i can see this log for this request in httpd/ 
access_log:

cent5n1.dom.local - - [13/Apr/2009:22:04:32 +0200] "POST /XMLRPC HTTP/ 
1.1" 200 4133 "-" "rhn.rpclib.py/$Revision: 102540 $"
192.168.1.12 - - [13/Apr/2009:22:04:34 +0200] "GET /XMLRPC/GET-REQ/ 
centos-5-base-x86_64/repodata/repomd.xml HTTP/1.1" 404 9 "-"  
"urlgrabber/3.1.0"

and in rhn/rhn_server_xmlrpc.log:

2009/04/13 22:04:33 +02:00 15832 192.168.1.12: rhnSQL/ 
driver_cx_Oracle.connect('Connecting to database', 'xe')
2009/04/13 22:04:34 +02:00 15832 192.168.1.12: xmlrpc/ 
up2date.listChannels(1000010042,)

and there is nothing in the cache.
What can be wrong there?

Thomas