Jean McCormack wrote: > Please update the copyright. > > line 993: eror > error
fixed. > > Unfortunately, I don't know enough about be with zones to be > able to say "yeah you've fixed this". The problem was that I was calling be_zone_supported() with the wrong thing. It should have been the zone dataset (for example rpool/zones/zone1) not the whole active zone root dataset (for example rpool/zones/zone1/ROOT/zbe-1). This worked in the November release but really shouldn't have. The fix for bug 4794 uncovered this bug which is this improper use of be_zone_supported(). Thanks, -evan > > Jean > > > > Evan Layton wrote: >> I need two reviewers for: >> >> 6122 be_activate can fail to promote the zone root dataset >> http://defect.opensolaris.org/bz/show_bug.cgi?id=6122 >> >> Webrev: >> http://cr.opensolaris.org/~evanl/6122/ >> >> >> I did not believe that this was a blocker bug until this weekend >> when I ran into a situation where I could not destroy a BE until >> after I had manually promoted the ZFS datasets for the zone's >> root dataset belonging to the active BE. This is not normally an >> issue and in most use cases people wouldn't hit this however if >> the chain of parents and clones gets confused enough it is possible >> that we can get into a situation where a BE can not be destroyed. >> The fact that it's possible to hit this destroy filure at all made >> this a blocker. >> >> There is a possible work around for this issue. However it requires >> that the user find the uuid of the BE and then compare that to the >> parentbe property of the zone datasets to find the correct zone >> dataset. Then that zone dataset and any subordinate datasets have >> to be manually promoted. >> >> Testing this required checking that the correct BE's zones root >> datasets were promoted when a BE is activated. Also checking that >> the correct snapshots belong to the correct zone datasets after >> the activation of the BE. >> >> Thanks, >> -evan >> _______________________________________________ >> caiman-discuss mailing list >> caiman-discuss at opensolaris.org >> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >
