GitLab

it.) Along with this, added a snippit of a Cisco config file that can
can be used to do this.

printed incorrectly.

in addition to the testbed VLAN id.

a :portnum argument. Makes it easier to arrange for testing my own
version of tmcd on boss by forcing all tmcc requests to a specific
port.

This replaces a couple of ad-hoc checks in the FreeBSD and Linux filesystem
code.  The idea is that we record a free range only if it is over a certain
length.  This has the effect of combining multiple short allocated ranges
into a single range, making for more efficient disk writes in the unzipper.

recognize OpenBSD reloc type

in the main group. Force them to be approved first!

waiting for Mike to work on Jail. This is an auditing module. Its
intended to serve two purposes. 1) Provide a common set of routines
for generating all that audit email from various scripts and 2)
provide a debugging hook for when things screw up via the web
interface and the user is too clueless to help us out, or the
information just got lost someplace.

The main function is:

	AuditStart($daemonize;$logname);

To start an audit, call AuditStart(). The first arg indicates if the caller
is wanting to daemonize. If not, just redirect stdout/stderr to a logfile,
and return. The logfile is optional; if not provided one will be created
based on the name of the script with mktemp. If the user wants to
daemonize, also fork and detach. This is provided as a convenience for
those scripts that tend to combine redirecting output and daemonizing. The
parent is exptected to exit, like any good mother of a daemon.

Okay, so all output is redirected to the log file. If a subscript is
invoked that also calls audit, that call is ignored, under the assumption
that the logging can be rolled into the parent, and besides it would create
a blizzard of email.

There is a package destructor (END) that is setup to email the log file to
the audit list, and if a log was created, the log goes to the logs
list. This is important. The audit list never gets any logs; it just gets a
two line record of what was done ("rmacct mike (by stoller)"). The log goes
separately to the logs list for inspection if needed at some point. This
makes the audit list very consise and easy to distill.

Like all good package destructors, you can tell if the script was
exiting with an error. If it was, then instead of sending the mail to
the logs list, send the message to tbops! The nice thing is that this
gets invoked no matter how you exit! No need to explicitly send the
email, unless of course you want it, but I have not written than
function yet!

Oh, for debugging. We can go stick in Audit calls when scripts
misbehave, and we can watch the output.

Cool, right? Really useful, right? Handy Dandy, right? Mike?

fix a bug in the free block recognition code for BSD

including the MBR (I Told Him So). The non-boot partitions are marked
NULL in the image descriptor, and so should be generally ignored,
although they will be saved and restored in the ndz file if the types
are not Unknown.

    Accept [switch.]<module>/<port> format for ports, so that we can
	deal with ports not in the database (mostly for my own
	debugging sanity.)
    A -n option that prevents assign from changing hardware settings
    	(though, unlike TESTMODE, does read some information from
	the switches)
    Private VLAN support, through the -x,-y, and -z switches. There
	are only 5 letters of the alphabet left, so I've given up on
	memnonic switches.
    Worked a bit on making VLAN deletion more efficient, but with
    	little sucess

Private VLANs work like so:
Make a primary private VLAN with:
snmpit -m myvlan-primary -y primary
Attach a community VLAN to it like so:
snmpit -m myvlan-community -y community -x myvlan-primary -z cisco2.1/15
Put some ports into the community VLAN:
snmpit -m myvlan-community pc1:0 pc2:0

dangerous to create private VLANs on switches with certain features
enabled (like the experimental net), so we'll use this column to
safeguard against any accidents.

one more TODO item

which needs to be renamed to something else.

used queries are now 0.01 seconds instead of 0.5 seconds!

    IP Port allocation:
    Low             High
    --------------- ------------
    30000           30255

    SSHD Port allocation ('ssh -p portnum'):
    ID              Port       SSH command
    --------------- ---------- ----------------------
    node1v-1        30000      ssh -p 30000 node1v-1.Tv2.testbed.emulab.net
    node2v-1        30000      ssh -p 30000 node2v-1.Tv2.testbed.emulab.net
    node1v-2        30001      ssh -p 30001 node1v-2.Tv2.testbed.emulab.net
    node2v-2        30001      ssh -p 30001 node2v-2.Tv2.testbed.emulab.net

and now sshdport). First off, take the DB out of the loop. There is no
reason to do DB locking since local nodes are not shared, and even if
they were, we now allocate port ranges to each experiment when it uses
virtual nodes (which is the only way to share a node anyway), so there
is never a case that port allocation between experiments has to be
coordinated. The only locking that is done is when the range is
allocated to the experiment, buts that done only once, and only if the
experiment uses virtual nodes (local or remote).

Also, now that its easy to create lots of jails on a single node, its
a good idea to give each one their own ssh port. This is recorded in
the DB nodes table (sshdport) and returned in the jail config from
tmcd.

totally forgot about (vnconfig does not return error status when a vn
device is busy). This was casuing the crashes.
Add a local only kludge to NFS mount /user and /proj into the jails. I
hate this. Should be using SFS, but without an agent thats not really
possible.
Add sshdport config setup from jailconfig. Look for that in the jail
config file and pass to jail.

Move bootvnodes to after ISUP so that the os_setup waits for the right
thing at the right time.

installed version in the new image).

passwords now have $ signs in them!

Clean up some stuff wrt scriptname and taint checking.

more than one jail on a node, in the same experiment. Will need a
similar field for the sfs server too ...

when eventsys is enabled, so the build was broken when events were
disabled (as they are in the testsuite).

change to expand copyrights when there is one to expand