[Spacewalk-list] Duplicate package provides

[Justin Sherrill]

Hey Greg,

I believe we just discovered this last week: 
https://bugzilla.redhat.com/show_bug.cgi?id=475894

The reporter (without digging too much) thought it might have something
to do with the fact that the packages had i686 and i586 rpms for the
same package.  But it does make a tad more sense now with your
explanation.  Thanks for looking into it!

We should be able to simply ignore that particular oracle exception, or
only insert the provides entry if one does not already exist (so we'd
only store it once).  That shouldn't be too much trouble.

-Justin


Greg Kuchyt wrote:
> I've been working with Spacewalk with a set of channels for Fedora 10.
> I was attempting to setup a kickstart tree when I ran into the output
> from the "lint" tool concerning a few packages that weren't present.
> Upon investigating it looks like I had six packages that weren't added
> to the base channel I setup. When I attempted to add these packages
> with rhnpush, I was received the following error for each package.
>
> Error pushing ./kernel-2.6.27.5-117.fc10.i686.rpm: Error 500Error
> Message:
>     ORA-00001: unique constraint (SPACEWALK.RHN_PKG_PROV_CID_PID_S_UQ)
> violated
> Error Class Code: 54
> Error Class Info:
>      Package Upload Failed due to uniqueness constraint violation.
>      Make sure the package does not have any duplicate dependencies or
>      does not already exists on the server (500)
>
> After doing some looking around in Oracle and at the packages I
> concluded that the unique constraint was being violated because the
> package provides the same capability more than once.
>
> [root at xxxxxx rpms]# rpm -qp --provides kernel-2.6.27.5-117.fc10.i686.rpm
> ->kernel = 2.6.27.5-117.fc10
> kernel-drm = 4.3.0
> kernel-drm-nouveau = 11
> kernel-i686 = 2.6.27.5-117.fc10
> kernel-modeset = 1
> kernel-uname-r = 2.6.27.5-117.fc10.i686
> linux-gate.so.1
> linux-gate.so.1(LINUX_2.5)
> ->kernel = 2.6.27.5-117.fc10
> kernel(x86-32) = 2.6.27.5-117.fc10
>
> I understand that this is an issue with the way the RPM was built, but
> I can't help wonder if there is a way to handle this situation within
> Spacewalk aside from failing with an Oracle index constraint violation?
>
> _______________________________________________
> Spacewalk-list mailing list
> Spacewalk-list at redhat.com
> https://www.redhat.com/mailman/listinfo/spacewalk-list