This is totally not cool. Running this tool modifies not just
agent.properties, but also just appends stuff to
/etc/libvirt/{qemu.conf,libvirtd.conf} and /etc/sysconfig/libvirtd. And
due to a disagreement between the changes to qemu.conf and
sysconfig/libvirtd, manages to break libvirtd. Nice! At n
> On Mon, Aug 13, 2012 at 12:59 PM, Marcus Sorensen
> wrote:
>> know what secondary storage we have, then cloudstack is canonical.
>> There is no concept in cloudstack of 'query and update myself about
>> which secondary storage devices the hosts know about'.
Fair enough. But it damn well better
In "Re: Review Request: Moved S3 and EC2 to DAO Implementation and removed
Hibernate dependency" and other recent threads, there seems to be quite a
bit of "we have to rip this tool out" or "we can't package CS with XYZ
because it's not ASF compatible. (see VMware, NetApp, et. al.)"
So instead of
> In basic mode, security groups are used for segregation but VMs are only
> allowed one network to live in. Why? Why couldn't an instance in basic
> mode belong to multiple networks?
No good reason. My comment about the ready availability of good-enough
hardware was not meant to imply that I su
Not only does the original suffer from re-computing the same value over
and over despite 'tmpfile' having been defined. This makes a maintenance
headache.
Worse, downloading files into /usr/lib* is VERBOTEN!! Options like
/var/{tmp,lib} are at least proper but it makes vastly more sense to put
the
