GitLab

widearea reloading support.

* New slot in the images table to store an access key which remote
  sites must provide in order to download an image (via https).

* tmcd returns a different kind of ADDRESS field from doloadinfo.
  Instead of the multicast stuff, return a URL that points to boss'
  web server. The URL is of the form:

   https://www.myemulab.net/spewimage.php?imageid=10013&access_key=abcdef

  which as you can see is fully specified; the client does not need
  to know anything else.

* New webpage and backend scripts appropriately called "spewimage"
  which also includes support for the http HEAD request (from wget) to
  avoid downloading images that are already on the node. I just
  learned about this HEAD request stuff today ... but otherwise these
  operate as expected, spewing the image if the access key is provided.

* Changes to rc.frisbee to deal with remote loading. In addition to
  URL support, I also added support for simple paths, the intent being
  that we will probably distribute images offline (say, at night) so
  that when a node reboots it doesn't actually have to wait 60 minutes
  for an image to download. I have not added any server side support
  for this yet though. Maybe later this week.

* Other bits and pieces and fixes to make this work.

Add functions for checking for wireless and subtypes
Get the proper link type into the top file in many places we had
  harcoded "ethernet"

printed out to indicates paths which do not obey the transitive property.

integrates some of the early RON support, adding a fake jail setup that
does a lot of what happens inside a jail, but without the actual jail.
Remains to be seen how well this is going to work out.

* Added the widearea node checkin stuff.

* Wrap up bootinfo in a subroutine that uses tmcd bootwhat on widearea nodes.

no hostname. Useful in the widearea where dynamic DHCP might not provide
a hostname.

* Add a bootwhat command that can be used instead of the bootinfo
  protocol, which is not appropriate for widearea cause its UDP. We
  lose the ability to have nodes "pxewait", but that is not actually
  necessary for widearea nodes since they are always allocated and up.

  Rather then duplicate code, I reorganized the bootinfo code so that
  I can link in the guts of it. There is some hackery to deal with the
  events that bootinfo sends, but so be it.

* Initial support for setting up accounts for the CMU widearea nodes,
  adding yet another way to specify accounts for a node. In this case,
  its a new node_type_attributes slot called "project_accounts" that
  is a comma separated list of group idx's. This is just a temporary
  measure to make sure that accounts are always set up on the nodes,
  the way they are for RON nodes. This will all need to change since
  we want to also create accounts on the fly for people creating
  experiments using the virtual nodes on each widearea node, but I
  have not figured that part out yet.

different then the original widearea code. Simpler, less dynamic.

First off, the wanodecreate script creates a new widearea_nodeinfo
entry.  These are nodes that will later checkin and be created as a
real node.  The input is a little xml file that you can use to specify
the stuff in the table entry (city, state, zip, etc). You can also
provide a privkey (no more then 64 chars), or one will be generated
for you.  For each one of these, create a Dongle Boot and stash the
privkey as /etc/emulab/emulab-privkey on the dongle. You do not assign
the IP address; the node will tell us that when it checks in.

A node checks in like this:

	https://$bossname/wanodecheckin.php?IP=$IP&privkey=$privkey
               &hostname=$hostname

The web page is simply a stub that makes sure the arguments don't have
any illegal characters, and then passes off to the backend.

The backend script checks the privkey and finds the widearea_nodeinfo.
The first time the node checks in, the node is created (db/Node.pm)
(nodes table, interfaces table, etc), and the node is moved to hwdown.
Subsequent checkins watch for changes to the IP or hostname, and issue
named_setup calls as needed.

all the delay agent events correctly - some more work is needed for clustering the paths
properly when transitive property of bottleneck links is violated.

bootwhat without code duplication.

place; the new wanodecheckin script that creates a widearea node when
it first appears.

experiment into which a node should be deposited after it is
created. Fall back to hwdown of course, when not specified.

plabslice, but still sorta behave like one.

Also put back some code that must have been removed at some point, to
initiate vnode setup on remote nodes (since we do not reboot widearea
nodes; they are are always allocated and up).

plabslice, but still sorta behave like one. Mostly fixing up some
special cases and using a different waittime calculation.

This will superceed the widearea_privkeys table that was used for
RON/Netbed but was never done well.

The privkey is just something we generate and store on the dongle (as
a simplification, we just create a dongle per node instead of
generating them on the fly) since there are not very many of these
widearea (CMU) nodes. We store the IP here for convenience (there is
an interfaces table entry of course).

The IP is updated when the node tells us it has changed, using the
privkey to uniquely identify it (for checkin).

physical type to pcvwa (our generic type for widearea virtual nodes).
Local node types are still mapped to pcvm.

the boot log since that is more likely to have something useful now.

servers cleanly. The thread library appears to hang up for about 30
seconds, sometimes a lot longer, waiting for something to happen that
does not (process stuck on a wait channel in the kernel).

doesn't actually have any.

reflect that dongle disk is hd0, real hard drive is thus hd1.

link type (assumed to be 'ethernet' for anything in the wires table),
also include types with the switch, type, and speed of the interface.

This allows us to ask for them more directly in the NS file.

one SQL injection hole, and shifted detection of probes  earlier in a lot
of other pages.  But some inputs that were marked PAGEARG_STRING should
actually be PAGEARG_ANYTHING, since they're text fields where quotes make
sense, and are escaped properly in the logic that handles them.

  approveproject.php3 - message
  editnodetype.php3 - newattribute_value
  newnodelog.php3 - log_entry
  newosid.php3 - description
  nodecontrol.php3 - startupcmd (node_control strips single-quotes from values.)

Increase max wait time for images based on 6GB at 1.6MB/sec (72 minutes)
Fix a couple of print statements.

that works for people who are just "average" ...