GitLab

This allows a number of constants to be set in the Cobalt config file.
We can also now set attributes like width and depth from qsub using
--attrs.  This also adds the hooks for using attrs=location:xxxyyy for a
resource reservation.

Cray likes to send and recive node id lists in a compact format.  This
allows cobalt to parse them.  This compact format is useful for display
as well.

I've gotten a test job to run end to end and relase resources.  ALPS
reservation set confirmed released, node added to pool, run and
released.  Lots of cleanup is needed.  Code is at prototype stage. Not
production ready, but a lot closer to it.

ALPS needs some extra work from the forker.  This is a modified
user-script-forker for ALPS use.  It doesn't inherit from
user-script-forker due to some differences in the informaiton passed.

user_script_forker should not be run on ALPS systems.

This adds in cleanup for ALPS reservations and resources.  Between the
signaling from the queue manager and release messages to ALPS, this
should retrun resources to the pool.  Testing pending.

Running of jobs should be testable from the system component.

This is needed to match the current cobalt interface that is expecting
information on the signaled process groups.  This gets passed up the
call chain by the system component's XMLRPC interface.

A modified version of the user script forker is needed so that we can
confirm the ALPS reservation and set the pg_id from the child.  This
will let apruns from a user script run.

Implementation of add_process_groups.  Testing pending
Most of the functionality for this has been pushed into the
process_manager and has been tested there, as well.

Adds find job location and reserve resources until, along with helper
functions.  Tests pending.  Draining and backfilling pending. Docs
pending.  Config options pending.

Testing is pending for this functionality.  This is the code that takes
queues, resources and reservations and determine which queues can
interact with eachother.

This extends nodelist for basic system status query.  Currently works
with alpssystem.

alpssystem can now be invoked as a system component.  It will initialize
query and update nodes.  This adds in the udpate behavior for nodes.
Persistence isn't supported yet and will be coming soon.  This is
currently using a base process manager.

Cray-specific node class added.  Cray node manager added.  Can
initialize from the ALPS bridge.  Core of alps bridge added.

I messed up something basic with pipes.  Added a second pipe, and set
proper O_NONBLOCK on the child stdout pipe.  This works correctly now.

This shortcircuts the /dev/null -> stdin override in the system script
forker so that string input works.

Allow forkers to take a string for stdin passed from the parent.  This
requires setting up appropriate pipes and fd rejiggering.  This appears
to handle cases that are larger than the pipe buffers appropriately.

The driver here is to support Cray's apbasil interface.

Conflicts:
	src/lib/DataTypes/ProcessGroup.py

Conflict resolved by use both.

Adds feature requested for #4320.  This will signal the head pid rather
than the process group.  Not recommended for use with csh scripts.

This is per Eric's suggestion in his commit comments.

Added in process management.  Working on tests now.

Finished test cases for basic functionality for ComputeNodes.

Without this, the tests fail to resolve the import resource directive.

Visibilty also corrected, so static checker works when including
'system' module.

Contains general resource base class, and some work on process group
handling.