GitLab

mode.

bad thing to do, and has recently stopped working. Use an index
ordering slot instead.

"tb-set-tarfiles" is like "tb-set-node-tarfiles" except that it
distributes the tarfile to all nodes rather than just one and that it
uses frisbee to distribute the file.

These changes involved 1) refactoring frisbee info from images table
into a new table, frisbee_blobs, 2) a new experiment_blobs table, and
3) a new tmcd command so the node knows how to get the files from the
server.

The changes where designed to be general purpose enough to eventually
support:
  1) Distributing arbitrary files (not just tarfiles) to nodes
  2) Perform arbitrary actions on those files
  3) Use arbitrary methods to get the files

As such the tmcd line is as follows:
  URL=* ACTION=*

where URL is currently:
  frisbee.mcast://<ADDR>/<FILE>
for example
  frisbee.mcast://234.16.184.192:18092/users/kevina/home-dir.tar.gz
and when we get around to using a master Frisbee server it could be
  frisbee://*
or it could be a file://, http://, etc.

and ACTION is currently:
  unpack:<LOCATION>
for example
  unpackt:/users
with future syntax to be determined.

use EmulabFeatures;

if (EmulabFeatures->FeatureEnabled("NewMapper", $user, $group, $experiment)) {
   # Do something
}
else {
   # Do something else.
}

where $user, $group, and $experiment is the current Emulab user, group, and
experiment the script is operating as. Any of them can be undef. Note that
features can easily be globally enabled or disabled (bypassing user/group
check). See below.

There are two scripts to deal with features. The easy one is the script to
grant (or revoke) feature usage to a particular user or group or experiment:

boss> wap grantfeature -u stoller NewMapper
boss> wap grantfeature -p geni NewMapper
boss> wap grantfeature -e geni,myexp NewMapper

Add -r to revoke the feature.

The other script is for managing features. To create a new feature:

boss> wap emulabfeature create NewFeature 'A pithy description'

which adds the feature to the emulab_features DB table. Use "delete"
to remove a feature from the DB.

You can globally enable and disable features for all users/groups (the
user/group checks are bypassed). Global disable overrides global
enable. There are actually two different flags. Lots of rope, I mean
flexibility.

boss> wap emulabfeature enable NewFeature 1
boss> wap emulabfeature enable NewFeature 0

boss> wap emulabfeature disable NewFeature 1
boss> wap emulabfeature disable NewFeature 0

To display a list of all features and associated settings:

boss> wap emulabfeature list

To show the details (including the users and groups) of a specific
feature:

boss> wap emulabfeature show NewFeature

Oh, if a test is made in the code for a feature, and that feature is
not in the emulab_features table (as might be the case on other
Emulab's), the feature is "disabled".

Add failureaction to virt_lan_lans so that we can allow lans to fail
during setup and still continue to swapin.

or more then one experiment, of course.

pairs. Used only by the pool daemon, to control concurrency between
the pool daemon and mapper.

new code. Note that this commit implements only the frontend part.

set n1 [$ns node]
set n2 [$ns node]
set n3 [$ns node]

set seg1 [$ns duplex-link $n1 $n2 * 0ms DropTail]
set seg2 [$ns duplex-link $n2 $n3 * 0ms DropTail]
tb-set-link-layer $seg1 1
tb-set-link-layer $seg2 1

set mypath [$ns make-path "seg1 seg2"]

set link0 [$ns duplex-link $n1 $n3 * 0ms DropTail]
$link0 implemented_by $mypath

sql/database-create.sql and database-fill.sql: add OF columns into virt_lans, add their descriptions into table_regex.

of desire to 64 chars.

devel tree).

an event and to actually fire those events during experiment swapout.

(class is switch or role is testswitch). Also add a node_type
attribute for all existing switch types: forwarding_protocols=ethernet

at different layers. Defaults to 2 in the DB.

cut down on the number of IDs we throw into the rspec advertisement.

snmpit on experiment swapin. Good debugging tool.

Add all the NS parsing and DB foo.

deleted, they still remain in the user table with a status of
"archived", but since all the queries in the system now use uid_idx
instead of uid, it is safe to reuse a uid since they are no longer
ambiguous. 

The reason for not deleting users from the users table is so that the
stats records can refer to the original record (who was that person
named "mike"). This is very handy and worth the additional effort it
has taken.

There is no way to ressurect a user, but it would not be hard to add.

happy.

sitevariables.

user is a local user. Instead, all users have to send along their keys
in the RedeemTicket() call, and those keys land in the new Emulab
table called nonlocal_user_pubkeys, and tmcd will use that table when
sending keys over local nodes.

This change removes the inconsistency in key handling between slivers
created locally and slivers created at a foreign CM.

(including an update to make schemacheck happy), and client-side script
modifications.

Tested by me on a few different machines. Ok by Mike.

the tb-set-node-os command with a second optional argument; if that is
present, the first arg is the child OS and the second is the parent OS.
We add some new features in ptopgen (OS-parentOSname-childOSname) based
off a new table that maps which child OSes can run on which parents, and
the right desires get added to match.  We setup the reloads in os_setup
along with the parents.  Also needed a new opmode, RELOAD-PCVM, to handle
all this.

For now, users only have to specify that their images can run on pcvms, a
special hack for which type the images can run on.  This makes sense in
general since there is no point conditionalizing childOS loading on
hardware type at the moment, but rather on parentOS.  Hopefully this stuff
wiill mostly work on shared nodes too, although we'll have to be more
aggressive on the client side garbage collecting old frisbee'd images for
long-lived shared hosts.

I only made these changes in libvtop, so assign_wrapper folks are left in
the dark.

Currently, the client side supports frisbee.  Only in openvz for now, and
this probably breaks libvnode_xen.pm.  Also in here are some openvz
improvements, like ability to sniff out which network is the public
control net, and which is the fake virtual control net.

* Add external resource identifiers to the reserved table so we can
  figure out the external node that is mapped onto the local (proxy)
  node.

* Change limit on fixed_iface to 128 chars to accomodate urns.

make a comment instead of 2.

'tables', 'reset', and 'synctables' commands. Rather than assuming all
VLANs live on the One True Experimental Net Stack (unless -S was given),
they all now check the database to see which stack they should be operating
on.

Changed how stack information is currently stored in the database. The
LAN attribute formerly known as 'stack' is now called 'class', and has
two valid values, 'Experimental' and 'Control'. The idea is to record
whether something is an experimental net or control net LAN, not which
specific stack it's on. The 'stack' column in the VLANs table now
records the actual stack_id on which the VLAN has been created.

This is all in support of testbeds that have multiple experimental-net
stacks, as Utah currently does.