GitLab

variable, which is now references in the toplevel GNUmakefile.in file.

link on the Begin Experiment page, remove the admin only test, and
make the palette bigger.

system, of the form:

	link0	fxp0  pipenumber1
	link0	fxp1  pipenumber2
	link1   fxp1  pipenumber

When dealing with trunks, since they may be EtherChannels, we have to
make another lookup on the switch to get the ifIndex for the whole
channel - can't just use the ifIndex of one of the ports like you
can from the Cisco CLI.

These changes make snmpit slightly slower - it now has to get more information
from the switches when it's going to create or delete VLANs. However, this
is on the order of fractions of seconds, so it shouldn't be too noticable.

to real VLAN numbers on the swtiches.

scheme that we discussed in email. Notifications and subscriptions now
take an "address_tuple" argument (I know, crappy name) that is a
structure that looks like this:

	char		*site;		/* Which Emulab site. God only */
	char		*expt;		/* Project and experiment IDs */
	char		*group;		/* User defined group of nodes */
	char		*host;		/* A specific host */
	char		*objtype;	/* LINK, TRAFGEN, etc ... */
        char		*objname;	/* link0, cbr0, cbr1, etc ... */
        char		*eventtype;	/* START, STOP, UP, DOWN, etc ... */

These can be a specific value, ADDRESSTUPLE_ANY if you are a
subscriber, or ADDRESSTUPLE_ALL if you are a producer. The reason for
the distinction is that you can optimize the match expression with the
extra bit of information, and the above structure can make for a
fairly lengthy match expression, which takes more time of course.
You should use address_tuple_alloc() and address_tuple_free() rather
than allocating them yourself. Note that host above is actually the
ipaddr of control interface. This turns out to be more convenient
since free nodes do not have virtual names.

Also added a new tbgen directly. This directory includes 3 programs in
the making:

tbmevd: Is the Testbed Master Event Daemon, to be run on boss and will
handle TBCONTROL events (reboot, reload, etc). It is just a shell of a
program right now, that takes the events but does not do anything
useful with them. Have not defined what the events are, and what DB
state will be modified.

tbmevc: Is the Testbed Master Event Client (akin to tmcc). It
generates TBCONTROL events which the tbmevd will pick up and do
something useful with. This program is intended to be wrapped by a
perl script that will ask the tmcd for the name of the boss (running
the event daemon).

sample-client: This is a little client to demonstrate how to connect
to the event system and use the address tuple to subscribe to events,
and then how to get information out of notifications.

Note that I have not created a proper build environment yet, so new
programs should probably go in the event dir for now, and link using
the same approach as in tbgen/GNUmakefile.in.

As much of it as possible switch-independant, and put in snmpit_lib.
Should work for arbitrarily complicated switch toplogies, as long as
there are not multiple trunks between two switches. (Multiple ports
combined into one trunk are fine, however.)

As a result of this, VLAN creations and deletions now need to operate
on all switches, not just on the ones that have ports in the VLAN.
This is because the traffic may have to traverse switches that have no
ports in the VLAN to reach other switches that do.

Not called yet. I've done simple testing, but need to do more, as this
could get us into major trouble if it has bugs.

all put together. Use --enable-events.

Also several changes for building on FreeBSD. I've left the older
Makefiles in place so Ian can continue to build in his environmant
(Linux?).

Fixed typo in API (s/event_notification/event_notification_t/).

Tell user (in README) to configure Elvin library with --enable-threads
to enable threaded API support.

Removed instructions on enabling server discovery feature (now optional).

Document event_register interface changes ("threaded" arguement, which
allows caller to choose between the thread-safe "threaded" API and the
simpler "sync" API).

Mention (in API) that event_notify and event_schedule don't deallocate
notification.

the threaded API; otherwise, use the sync API.  This option allows
multi-threaded programs to access the threaded API, which is thread-
safe.

when backing up (cause of an error that needs to be fixed) since not
all browsers handle this the same. Instead, redraw the form with all
of the original info and a list of error messages at the top.
Conceptually simple change, but it turns out to be a pain to implement
since you need to combine the form and processing code in one page
(well, its just a lot easier to do that), and then change all of the
forms to deal with a "default" value. That is, each different kind of
input tag (text, radio, select, checkbox, etc.) requires slightly
different changes to do that. Lots of forms, lots of entries on the
forms, and its a long slow tedious process. Much nicer though, although
the code is a bit harder to grok. At the same time, I added a lot more
sanity checks of the information being passed in.

The other change is to deal with how browsers handle the back button
on a form thats been properly submitted. Not all browsers use
the cache directives the same, and I was often typing back, only to
have some form get reposted. Thats a major pain in the butt. The way
to deal with that is to have the processor send out a Location header,
which modifies the browser history so that the post is no longer in
the history. You back up straight to the unposted form (if its in the
cache). I've done this to only some forms, since its a bit of a pain
to rework things so that you can jump ahead to a page that spits out
the requisite warm fuzzies for the specific operation just completed.

I've done newproject, joinproject, update user info, newimageid, and
newimaged_dz forms.

emulab address while I'm at it.

an infinite loop rather easily via the NS file TCL hooks. Added a
perl wrapper around parse.tcl called parse-ns, which forks a child to
run the parser. The parser is invoked "nice +10" and the CPU limit for
the child is set to 60 seconds, which should be enough for any parse.
If the limit is exceeded, send email to tbops since this indicates a
big problem or a user being dumb/malicious.

installed.

Second version.

the defs file. Re-arranged some of the existing snippets to make it
somewhat more cohesive, though it could still use a lot more work in
this area. Also, moved all of the remaining stuff about what to copy
from an old boss node (which was the original purpose of the document)
to the end, so that it's more geared toward people setting up their
own boss.

Still needs some documentation (and probably a script for auto-generating)
DNS config files.

consistent 0ms when the intent is no delay, and add an explantion of
what 0ms means (no darling, it doesn't mean that packets travel
through an ethernet wormhole).