Description of problem: After doing a fresh install on to existing software (md) raid arrays and rebooting, the arrays are all degraded and only have one member listed. Once I fixed them up with mdadm they stayed that way accross reboots. Version-Release number of selected component (if applicable): Rawhide from April 27 and test4 both showed this problem, but I think it goes back a ways and I didn't notice until recently. How reproducible: This seems to be happening all the time, but my sample size isn't that big. Steps to Reproduce: 1. Do a fresh install on existing raid arrays. 2. Check the status of the arrays after rebooting. 3. Actual results: Raid arrays degraded Expected results: Raid arrays functioning with the setup defined during the install. Additional info: I didn't test this with arrays assembled during the install and don't know if that would make a difference or not.
I just did an install with today's rawhide + the changes to use mdadm in anaconda instead of the old raid bits and can't reproduce this. I _suspect_ this was due to horkiness with raidautorun that won't get hit now. Please reopen if you see with tomorrow's rawhide or later