On Wed, 2007-03-28 at 10:09 -0500, Michael E Brown wrote: > On Wed, Mar 28, 2007 at 01:44:35AM -0400, seth vidal wrote: > > On Tue, 2007-03-27 at 17:01 -0500, Michael E Brown wrote: > > > 1) Does anybody have any objections to doing it this way? Is this the > > > best way to do this? > > > > > > 2) How do I supress the urlgrabber error output? I still get the > > > following error output even though everything appears to complete > > > successfully: > > > > > Thanks for the feedback, Seth. I really appreciate it. > > Did you have any feedback for the first part, ie. setting yumvars? > > > > > How would the use distinguish b/t this repo doesn't exist and the site > > is down? How would yum? Why is having it return a 404 and not just an > > empty repo better? > > Excellent points to which I dont have any really good answers. I couldnt > think of any bad consequences to having the user run without the > hw-specific repo if there were, eg. a spurious 404 error. > > > the url above could easily be a cgi with a redirect in it. If your > > At the point where I coded this, I didnt believe that I had the > permissions to run CGI on our system. Turns out I was wrong. I've > whipped up a quick shell script that redirects to an empty repo for > unknown system ids. My only problem now is that I am running an actual > CGI, so this is going to scale like crap. My server-side is a 2x1Ghz > PIII system - a 5 year old Dell PowerEdge 2450. It is going to run out > of steam quickly running cgi's. (and no, I dont have admin access to the > box.) > > > systemid has a repo, it sends you there, if not it sends you to an empty > > repo. > > I suppose another concern would be people who want to mirror my repo but > cant run cgi scripts. >
then try this: 1. put aliases in for apache 2. have the 404 errordocument be a redirect to the empty repo. -sv _______________________________________________ Yum-devel mailing list [email protected] https://lists.dulug.duke.edu/mailman/listinfo/yum-devel
