From: Dan L. <da...@la...> - 2007-08-28 01:44:24
|
On 28 Aug 2007 at 1:12, Peter Buschman wrote: > > Dan: > > I have daily regression tests working for Solaris 10 (sparc) but have > kept the output off of bacula-beta for the moment to avoid spamming > the list. At the moment, my script does both the all-non-root-tests > and all-non-root-tape-tests using the current SVN head every night > beginning at midnight CET. My script is a bit tailored to my > installation, but if you have thoughts on how to best build a shared > system for multiple platforms, I'm game.* My initial thoughts: http://dan.langille.org/2007/08/19/a-multi-platform-regression- testing-machine/ My conclusion: a multi boot system. There are ways to specify, at shutdown time, what the next OS to boot should be. I'm sure that can be done. > I suspect the enabling factor is less in having the same system > perform the tests through multiple boot cycles than in allowing > collaborative testing through multiple hosts (both physical and virtual). I could go virtual, but I getting access to the real tape drive is harder. I' not sure yet, but perhaps virtual without tape drives is the way to go. -- Dan Langille - http://www.langille.org/ Available for hire: http://www.freebsddiary.org/dan_langille.php |