type in the Utah Emulab database.

/usr/testbed/images

tunnels between containers and physical nodes since we use gre keys on
containers. Since we cannot stop using containers, lets just bite the
bullet and start making all tunnels use gre keys.

does appear to be failing at some sites.

And fix netmask handling during host boot.

node type, except in Utah, and this was causing odd errors
on the InstaGeni racks. Curious ...

node type, except in Utah, and this was causing odd errors
on the InstaGeni racks. Curious ...

Don't consider iSCSI-created block devices as local disks for local block
stores. Allow iSCSI devices to have filesystems created and mounted
(though there are still some ugly issues at reboot time). Some cleanup.

Also come cleanup/refactoring.  The parser will now enforce a variety
of things (to be documented) when a user constructs a blockstore and
attaches it to either a node or a link/lan.

Also, refactored the OS specific stuff into a new liblocstorage.pm
which is included by rc.stoarge. Note that there is no OS-independent
libstorage.pm, at least for now.

Supports the three coarse-grained placements we decided on:

  "SYSVOL" is special. You can declare a single blockstore with this
       placement and it will create a "native" (ufs/ext) filesystem on
       the 4th partition of the boot disk. This is how you create an
       extra storage partition that can be captured in a custom image.
       We don't use a volume manager here because imagezip doesn't
       recognize any of them (lvm, zfs, vinum).

  "ANY" coalesces all "available" storage from all disks into a logical
        volume manager pool and dishes out storage from that for
        individual blockstores. Typically this would include, the 4th
    partition of the boot disk (if not in use) and the second hard
    drive. If the machine has more than 2 drives, it will include
    all the extra drives.

  "NONSYSVOL" coalesces all "available" storage that is NOT on the
       boot disk into a logical volume manager pool and dishes out
       storage from that for individual blockstores. This case is if
       you want to avoid interfere with the system disk.

Only implemented on FreeBSD 8/9 with "vinum" right now. It only creates
"concat" (JBOD) volumes right now.

This stuff will probably get split out into its own perl module(s) at
some point, as it is getting large.

Next up is LVM on Linux and then maybe ZFS on Freebsd.

Mike unearthed another round of things we need to check to keep users
from shooting themselves in the foot too readily.  Made it through
a fairly complete set of input tests and came up with a couple of additional
checks myself.

Move some of the hacky duplicate functionality code for blockstore objects
into a finalize() method, called by sim.tcl's run() method.  This code
does last minute validity checks and assignments (e.g., putting the disk
space desire onto node objects).  Added the code here to check for
overlapping mount points where blockstores are attached to real nodes
(local storage).

This reverts commit 102a123e.