Hi all,
I'm running into an issue related with updating an existing patch in a ISS slave which has been modified in the ISS master.

1 - I create and publish a custom patch in the ISS master.
2 - I sync the patch in the ISS slave. A new patch is created in the slave.
3 - Now, I update the patch in the ISS master adding a new related package to the patch. Patch is now updated in the master containing the new added package. 4 - After syncing the slave using "--force-all-errata", the new package is added to the patch in the ISS slave (which is good and expected).

Here comes the issue:
5 - I update the patch by removing a package from the patch in the ISS master. The patch is updated in the master and now the package is not showed anymore as related with this patch. 6 - After syncing the slave using "--force-all-errata", the package is NOT removed from the patch in the ISS slave. So, related packages for the given patch are not the same between ISS master and slave after synchronization.

I took a look at how the erratas are handled during syncing and found out that there's a specific code to prevent the cleanup of "rhnErrataPackage" after processing the errata (called from "processErrata"):


If I remove this conditional statement, then "rhnErrataPackage" is processed and not excluded from cleanup. So the package is going to be removed and patches will be successfully synchronized.

Problem is this code is present since the initial commit, so my questions are:

Do you know why "rhnErrataPackage" is explicitly excluded here from cleanup?
Should we get rid of itat this point in time or there are other implications about removing it?

Any feedback or feelings are more than welcome!
Many thanks.


Pablo Suárez Hernández - <psuarezhernan...@suse.com>

Spacewalk-devel mailing list

Reply via email to