System errata count - force update

Latest response

Hi! Satellite 5.6 I've done a yum update on a system registered in Satellite. Not a scheduled update from the console,
Before the update system had let say 27 erratas. After the yum update, every package is at its latest.
I'm trying figure out how force the update at the console. It still shows the 27 erratas.

-I've tried cleaning the local cache repo
-yum clean
-rhn_check
-rhn-profile-sync

Still showing the same 27 erratas to apply.

SA
Log in to join the conversation

Responses

R. Hinton's picture Guru 6827 points
12 January 2015 5:13 PM

Not sure if this will apply for you - but I had the same issue on my RHEL 6 Satellite 5.6 server until I performed this fix, which says to update to a newer set of spacewalk-java rpms that were not part of any ISO channel I downloaded from Red Hat, and my satellite servers missed about 13-14 updates to spacewalk java because they are not facing the public internet see discussion here.
- link to spacewalk-java
- link to spacewalk-java-config
- link to spacewalk-java-lib
- link to spacewalk-java-postgresql
I didn't initially include spacewalk-java-oracle, and if you happen to have used oracle (not typical, but it can be done) with your satellite 5.6, then get the spacewalk-java-oracle update too.

I also found a brief problem with rhncfg rhncfg-client and rhncfg-actions that I had to downgrade temporarily until I received a newer edition because I could not ingest or deploy configuration files https://access.redhat.com/solutions/1250053,

Hope this helps, if it does not, open a case with Red Hat, have a copy of your relevant logs, or the sosreport if needed (and if you can share it)...

SA Community Member 27 points
12 January 2015 6:36 PM

All my spacewalk-java are at the latest for 5.6.

R. Hinton's picture Guru 6827 points
12 January 2015 9:15 PM

Contact Red Hat Support...

When I've faced this, the problem has often been that taskomatic may have failed in building the repo properly. I've had this where the satellite server believes one thing, and the client system that is attached can not update regardless of what the satellite says.

I'd recommend looking through the tomcat and rhn logs to look for errors, and also look for failed events in the web gui..... Personally, I was receiving out of memory issues and I made adjustments with the configs listed in that link.

Log into your satellite server web gui as an RHN admin, look under the "admin" tab (top, left, between "users" and "help", and if it is not there, you're not logged in as an RHN admin)... Then click on "Task Schedules in the grey field at left, then look through failed events between the "Channel-repodata-bunch" and other areas..

I'd look at the tomcat logs via command line, but you can also see them in the web gui near the same area in the above paragraph... Hope this goes well for you - I'd recommend contacting Red Hat support.

Cameron Kerr's picture Active Contributor 178 points
20 July 2015 3:05 AM

I'm finding that I need to do the following:

  • I use the 'Remote Command' feature to run my update sequence, cleanup, reboot etc, which I do with the help of System Groups.
  • As part of that, I include rhn-profile-sync in the sequence of commands that I run, although I don't know if that is necessary (I suspect its not; I'll have to try it next patch cycle; certainly my experience is that its not enough to do just that)
  • The errata counts will not be recalculated immediately, but appears to be done in a schedule job; there is a scheduled task for that, as shown below.
From /var/log/rhn/rhn_taskomatic_daemon.log:

... 2015-07-20 11:30:00,035 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 1
... 2015-07-20 12:00:00,042 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 1
... 2015-07-20 13:00:00,040 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 1

    The following four lines are apparently emitted by one run (updates every 10 minutes).
    Note that I hit Package List Refresh button on somewhere between 200-300 machines at 2:13, and on
    a few other things before that.

... 2015-07-20 13:50:00,018 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 3
... 2015-07-20 14:00:00,033 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 3
... 2015-07-20 14:10:00,027 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 19
... 2015-07-20 14:20:00,035 ... INFO  com.redhat.rhn.taskomatic.task.ErrataCacheTask - In the queue: 264

If you login as 'satadmin' (of the top-level Org), you can see in the Admin > Task Schedules > errata-cache-bunch the following description:

    "Recalculates errata cache for a given server or channel"

Ah, that explains it. This job gets run every 10 minutes.

What I think I'm going to be doing is after I apply my update sequence on a group of machines, I'll wait until it finishes, and then I'll schedule a package list refresh on each of the machines [or rather, on the group]. Then, 10 minutes later, it should reflect in the interface. Alternatively, if I'm impatient (eg. last group of machines before heading home for the night), I might log into the Satellite as the Satellite Administrator, and perform a Single Run Schedule on the errata-cache-bunch, which it seems may take on the order of 10 minutes for perhaps 50-100 machines anyway.

This is for Satellite 5.6 (and likely prior), with both RHEL 5 and RHEL 6 managed servers.

I'm not sure why it seems to be necessary to schedule a Package List Refresh and why this isn't done automatically....

Cheers,
Cameron

Close

Welcome! Check out the Getting Started with Red Hat page for quick tours and guides for common tasks.