Description of problem: source package rhn-virtualization contains: rhn-virtualization-common rhn-virtualization-guest rhn-virtualization-host Version-Release number of selected component (if applicable): rhn-virtualization-5.3.0-13.el5sat How reproducible: always Steps to Reproduce: 1. run tps test with errata for rhn-virtualization-common Actual results: rhn-virtualization-guests conflicts with rhn-virtualization-host, both package are built by one source. Expected results: if there is two separated source packages test automation would be very easy and save time for QA. Additional info: it was discovered by TPS test TPS-RPM fails in installing all packages in errata to stable system. TPS-RHNQA: failes in: TPSRESULT: Binary Package Completeness Returning: FAIL Binary package install result: FAIL Error: rhn-virtualization-host conflicts with rhn-virtualization-guest Error: rhn-virtualization-guest conflicts with rhn-virtualization-host
It's somewhat difficult to automate the testing of an erratum when any of the packages conflict with one another. For example, imagine a system that has none of the RPMs on. Which do you install first? If they're 2 separate updates, the combinations are easy to test. The other advantage of a split is having a smaller update if some issue arises that only affects -guest or -host. Smaller updates make it easier on everyone.
This bug report lead to a deeper investigation of the problem leading to the following: rhn-virtualization-guest subpackage was created during development phase of Satellite 5.0. Later in the development cycle of Satellite 5.0, there came a decision not to use rhn-virtualization-guest subpackage, as rhn-virtualization-host & rhn-virtualization-common packages installed on a virtualization host will do the intended job just fine. For some reason, this -guest subpackage has never been dropped from the .spec file. The errata leading to this bug report was historically the very first rhn client tools errata which went out through ET and as such got introduced to its friends (TPS among them). The true fix for this problem is in fact to drop the -guest subpackage from rhn-virtualization. Here I'd like to take the chance to change the summary of this bug report to: "Drop -guest subpackage from rhn-virtualization" spacewalk.git master: 4cc2533d345d9c3261344357e34b5db7ea969e27 e99c7f98dc08aa839742beea0624005833de2030 79a301bcaa2ce83127a5a65c07b14e1b07e98d36 2df4b968155893179416e15ae917050f8920649f satellite.git SATELLITE-5.3: b56ed42c5177e07f6343f5b7030fbe42e49778a9 3d3bc2841533a9c0db2baa0fa0842945446448fd 18d0b323f188090f9d131e05f822daa8d80b9a28 01a20fa8890829d8c21c5d351bc63dd0d38ca32f tagged & built rhn-virtualization-5.3.0-14
verified testing procedure: * rhn-virtualization-guests is not built * tps tests passed https://errata.devel.redhat.com/tps/errata_results/11934 results: TPSRESULT: Rebuild Package Count Test-selinux Returning: PASS PASSED: SRPM Rebuild Test finished successfully. The following packages were built: /usr/src/redhat/RPMS/noarch/rhn-virtualization-common-5.3.0-14.el5sat.noarch.rpm /usr/src/redhat/RPMS/noarch/rhn-virtualization-host-5.3.0-14.el5sat.noarch.rpm List of Build Logs: rhn-virtualization-5.3.0-14.el5sat.src.rpm-noarch-rebuild.log.gz
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2009-1660.html