Open Stack

Hey Y'all,

In my seemingly never-ending quest to nail down OpenStack, I'm having 
two odd issues with Horizon I was hoping someone could shed some light 
on.  I'm using RDO OpenStack Icehouse RC1.

1: My test user has been creating and destroying VM instances for a 
while for various tests.  It seems, however, that the Horizon dashboard 
that shows how many CPUs, RAM, instances, etc on the front page is 
flawed, as right now there are no running instances at all, but it still 
shows me running 1 CPU, 1 instance and two GB of RAM.  It's like an 
instance got destroyed but whatever is watching the utilization metrics 
didn't get updated.  Is that a known issue?

2: It seems that every once in a while my session in Horizon times out 
(which is expected), and then I try to log in again, and it flips to a 
page saying something to the effect of "Something went wrong!  Refresh 
your page and see if it helps".  It doesn't help.  It seems the only 
thing that helps is restarting keystone, then I can log in and 
everything looks fine.  Horizon is using memchached, if that makes any 
difference.  It's like it's caching old/bad credentials or something and 
confusing Horizon.  Does that ring any bells?

Thanks again for any hints!

-erich