used in the database, for backward compatibility.

the nastier shell scripts I wrote to somewhat cleaner perl.

stderr separately to avoid messing up the XML output of the command
due to warning or the like.

which I did not expect. Not sure I like it ...

it from the checkout, and make gets confused about whether config.o
should be made from config.cc or config.c . Add a hacky rule to the
makefile to build config.o from config.cc .

openvz version.  This deals with netlink and pkt sched changes.

values for configuration parameters, such as some of the cooling
parameters, scoring values, etc.

Also add a -D option to dump these parametrs, as well as the command
line, for posterity.

Refresh the dependency list at the end of the makefile

add escaping to the table_regex command.

(with quotes)

into libtblog.

and some lint-ing

problems.  When someone other than we does an install.

library at the moment. Maybe that library should export a boottime
method?

type counts for dynamic pclasses, which (correctly) never got added to
those counts in the first place.

Explicitly mark dynamically-created pclasses as such, so that we can
treat them specially in a few circumstances.

is rebooting. Just when creating.

the root fs of the vnode before it is booted. Did not want to
duplicate this code in the libraries.

mountd sillyness.

are tired of typing.

enable/disable on each port (one XMLRPC per). Instead, do this as part
of the remotedovlans and remotedestroyvlans.

Some reorg in doVlansFromTables() do deal with vlans that need to be
trunked in dual mode. That vlan must be created first, the ports
trunked, and the rest of the vlans created.

information about whether the interface needs to trunk in dual mode
instead of equal mode. Normally we use equal, but the supervlan
created between shared hosts needs to be trunked in dual mode so that
"raw" packets are delivered (or else non-encap packets go into the bit
bucket).

When tearing down vlans we do not use individual trunk disables, but
do it one block from destroyvlans. This is way way faster then an
XMLRPC call for each individual port.

For trunk ports, no longer use the vinterfaces table, but instead use
the trunk slot of the interfaces table. I removed a Keith Sklower
comment that said how silly that was; he was right, and now its
fixed.

changes, and we want old elabinelab to continue to function. This
commit is for the original code.

information about whether the interface needs to trunk in dual mode
instead of equal mode. Normally we use equal, but the supervlan
created between shared hosts needs to be trunked in dual mode so that
"raw" packets are delivered (or else non-encap packets go into the bit
bucket).

instead of a join through the vinterfaces table, use new "trunk" slot
of the interfaces table, which is set accordingly by the mapper
wrapper.

cluster nodes. Not going to try and describe all these changes.

Note that I have not back ported this into the old assign wrapper. We
move inexorably forward.

Worth mentioning:

* Users get VMs only on shared hosts.

* Multiple experiments from multiple projects can share a node.

* Nodes that are acting as shared hosts are in a holding experiment
  and have a tag in the reserved table. All of the links in the
  experiment are tied together in one giant super vlan. We then
  multipleax over that using our standard mechanisms (veths, vlans,
  whatever).

* Lots of complication in the link setup code for dealing with links
  between a virtual node on a shared node, and a private physical
  node. Requires additional vlans and trunking between those
  interfaces. To make life easier, all of the links in the afore
  mentioned super vlan are trunked in dual mode.

* I had to change Mike's code that does the encap determination, since
  openvz nodes cannot do veth encap. I add some new osfeatures
  (veth-ne, veth-en, and vlans).

* On and on and on ...

Other minor bug fixes.

to it. The VMs then get bind mounts of the actual user and proj
directorys that are approriate for that VM. Seems like a small change,
so why did so many lines change?

Oh, there is some obvious danger here.

trunk mode. This replaces what we currently do in snmpit, which is to
use the vinterfaces table in an ugly query.