summaryrefslogtreecommitdiffstats
path: root/genome-bootstrap
Commit message (Collapse)AuthorAgeFilesLines
* Fixing genome-bootstrap licenseBrenton Leanhardt2008-07-221-2/+4
| | | | | | https://fedorahosted.org/genome/ticket/15 The Red Hat Legal team has approved the GPLv2 license for all Genome tools.
* Other then the license issue in ticket #15, rpmlint is now silent for ↵Jeroen van Meeuwen (Fedora Unity)2008-07-222-5/+11
| | | | genome-bootstrap
* Increasing the number of tries genome-bootstrap asks for the IPBrenton Leanhardt2008-07-212-2/+2
|
* Updating the versionMatt Hicks2008-07-161-1/+1
|
* More updates to the cloudmaster and genome-bootstrap to record ip addressMatt Hicks2008-07-161-2/+2
|
* Updating genome-bootstrap and cloudmaster to store the repo and ipaddress in ↵Matthew Hicks2008-07-162-8/+9
| | | | the cloud
* Changing build system over to Koji. Now the MakefilesMatt Hicks2008-07-121-1/+1
| | | | | will only generate src RPMs. Koji builds can then be submitted to specific distribution builds.
* Minor bug fix for genome-bootstrap discovered by the django-exampleBrenton Leanhardt2008-07-112-7/+8
| | | | | | | | | | | | | | The django-example only has one class. A String was getting returned where an Array was assumed. Yay for no unit tests for this. On a side note, the django-example was useful for uncovering a few corner cases with how we use Restr. The problem with Restr happens whenever 1 of something was returned. While it's nice enough wrap things in an Array if multiple items are returned, since Restr's XmlSimple instance is not setting 'forcearray => true' we have to wrap the call. It would probably be a good idea to ask the Restr maintainer about the 'forcearray' parameter.
* Revved the Release for genome-bootstrapGreg Blomquist2008-07-101-1/+1
|
* Merge branch 'master' of git://oss1-repo.usersys.redhat.com/pub/git/toolsGreg Blomquist2008-07-103-51/+32
|\
| * genome-bootstrap now calls cobbler directly to 'get_status'Brenton Leanhardt2008-07-103-51/+32
| |
* | Changes to genome-bootstrap:Greg Blomquist2008-07-102-19/+24
|/ | | | | | | | | | | | | | | | | | | | | | | | | | * https://engineering.redhat.com/trac/IT_Architecture/ticket/209 - genome-bootstrap now filters out any cobbler profiles with "__Appliance" in the name during default mode. This will reduce confusion when users are selecting the cobbler profile. NOTE!: The cobbler profiles have been updated on oss1-repo.usersys.redhat.com but if there are any cobbler profiles in a git repo that get laid down with a new GenomeAppliance machine, these will have to be updated too. Here's how the profiles were updated: GenomeCloud* -> Cloud__Appliance* GenomeRepo* -> Genome__Appliance* GenomeRepoGuest* -> GenomeGuest__Appliance* * https://engineering.redhat.com/trac/IT_Architecture/ticket/210 - The machine_types.rb file on oss1-repo.usersys.redhat.com was updated to no longer ask for the "repoappliance" fact value. Now, genome-bootstrap's default mode assigns the "repoappliance" parameter the value of the genome repo. * https://engineering.redhat.com/trac/IT_Architecture/ticket/211 - genome-bootstrap no longer asks for "additional facts" in the default mode. The assumption is that if you want to provide additional facts you would use the advanced mode.
* Proxying the cobbler requests through apacheBrenton Leanhardt2008-07-092-2/+2
| | | | This keeps us from needed our hack to bind the rw api on all interfaces
* Fixed a couple of bugs in genome-bootstrap:Greg Blomquist2008-07-092-11/+30
| | | | | | | | | | | | | | | | | | | * replaced references to "fqdn" with "system-name" in the run method. - Instead of collecting the fqdn, or constructing based on Red Had-specific DNS information, we now only collect enough information for a cobbler system name. * eliminated ability to specify a genome repo by "leaving off the -repo" - Since we cannot assume that any repo (or any machine) will have a "usersys.redhat.com" suffix, the ability to specify a "partial" repo name has been removed. Users are now prompted for the fully qualified domain name of the repo. * selecting first element from installed_systems - With the introduction of "restr_get" that makes sure that all RESTR GET calls return an array (even of only one element), the "get_installed_systems" method needed to change to select the first (and only) element in the installed systems array. Otherwise, "cannot convert Integer to String".
* Renaming everying everest to genomeChris Alfonso2008-07-0814-0/+1087