GitLab

This was breaking pclasses badly.

In the process, added a few new flags for debugging pclasses,
DEBUG_PCLASS_MORE, and DEBUG_PCLASS_TONS.

event_arg_get().

Add some checks for 'free' nodes that are not allocatable.

	* db/audit.in: Include the list of nodes that are not reserved but
	have an eventstate that makes them unallocatable.

	* www/dbdefs.php3.in: Add POWEROFF and ALWAYSUP node states.

	* www/nodecontrol_list.php3: Add an asterisk next to the free
	count for type(s) that have free, but unallocatable nodes.

	* www/shownodetype.php3: If a node is free, but unallocatable, put
	a yellow ball next to its name instead of a green one.

We were exporting to ops because of the opsnode experiment and
this breaks the ever-fragile BSD exports semantics when the fs node
is not the ops node.

Also got rid of another shark reference!

size on FBSD4 machines.  Didn't affect the frisbee MFS, cuz we are now
using FBSD5 (and in the FBSD4 MFS, growdisk hasn't been reinstalled in ages).

Bring linktest in line with the new event system world order and do some
generic cleanup.  I'm pretty sure I broke backwards compatibility with
old disk images though...

	* configure, configure.in: Remove linktest.h.in from the list of
	template files.

	* event/linktest/GNUmakefile.in: Don't make subdirectories for
	executables since it can mess up the dependencies.  Don't install
	tb_compat.tcl stuff anymore.  Check for '/usr/local/bin/pathrate'
	and friends, and print out a warning if they are not found.  Make
	a version file for the linktest daemon.

	* event/linktest/linktest.c: Only subscribe to the events we care
	about 'STOP/KILL'.  Subscribe to TIME START events and kill any
	linktests that are in progress.  Send events through the local
	elvind instead of running tevc.

	* event/linktest/linktest.pl.in: Don't run the modified NS anymore
	to generate the topology file, use gentopofile to do that.  Send
	COMPLETE events to indicate that the tests were finished instead
	of STOP so it works with event-sequences.  Use emulab-sync to
	report errors between linktest daemons.  In addition to sending
	REPORT events, send LOG events to the SIMULATOR agent so the
	messages end up in the report e-mail.  Fix wait_all so it collects
	child processes immediately and doesn't leave zombies lying
	around.  Make the static route test work with topologies that
	aren't fully connected.  Change ownership of the linktest error
	files from root to the swapper.  Bump latency tolerance up to
	2.5ms.

	* event/linktest/ltevent.c: Use event_schedule instead of
	event_notify so the events go through the scheduler.

	* event/linktest/run_linktest.pl.in: Wait for COMPLETE events
	insetad of STOPs.

	* event/linktest/linktest.h.in: Removed.

	* event/sched/error-record.c: Include linktest error output in the
	report e-mail.

	* event/sched/event-sched.c: Linktest now sends COMPLETE events
	for START events.

	* tbsetup/gentopofile.in, tmcd/common/config/rc.topomap: Piggy
	back the generation and management of the linktest config file
	onto the generation of the topomap.

	* tbsetup/ns2ir/sim.tcl.in: Add "linktest" event to the Simulator
	object that runs linktest at level 3 (latency, static routing, and
	loss).

	* tmcd/common/libsetup.pm: Add TMLTMAP constant that contains the
	path to the node local linktest config file.

	* tmcd/common/rc.linktest: Pass '-u' option to the linktest daemon
	so it knows what user should to change ownership of files to.

	* tmcd/freebsd/jail/mkjail.pl: Copy the linktest config file
	(ltmap) into the jail.

frisbee MFS, and stated will send an apod out to it. This updates the
previous revision in which I was doing this for RELOADDONE state, so
that we maintain backwards compatability with older frisbee MFSs.

for the server to reboot us.  This avoids a race condition that occurs
under load (see the comment in the code).

the 'one-true-awk' version of awk, botches the calculations when the
values get close to MAXINT.  Instead, just create a hex string since
sysctl knows about hex numbers.

the "race" problem.

turned on.

project only. This avoids leaking out group member names.

There is also some unrelated fixup code in this revision. Ignore it.

Okay, I implemented a primitive Knowledge Base! The current contents are
*all* the existing FAQ entries, which I entered manually. Here are the
details.

* My reason for doing this is that we need something very simple. The wiki
  is too much of a barrier, and its search capabilities are pathetic.

* The search page for the Knowledge Base is:

	https://www.emulab.net/kb-search.php3

  Fairly primitive keyword search. Turns out that mysql 4.0 has a bunch for
  really good text searching functions built in, but we run 3.23 ... so I
  had to roll it myself. So, its a simple keyword (space or comma
  separated) search, no regular expressions.

* Each DB record has a "faq_entry" flag, so creating the current FAQ on the
  fly from the DB is easy. See:

	https://www.emulab.net/kb-faq.php3

* In reddot mode, you can add new KB entries:

	https://www.emulab.net/kb-manage.php3

  The form is fairly obvious but here are details anyway:

    Section Name: Choose an existing title, or make up a new one.
    Title:        The title of the KB (or FAQ) entry.
    Faq Entry:    Check this box if the new entry should show up in the FAQ.
    X Ref Tag:    A token so you can refer to other KB entries by name,
                  instead of by its index. Within the KB entry you would
                  write: <a href=kb-show.php3?xref_tag=sometag>
    Body:         Whatever you like. I took the existing FAQ entries and
                  stuck them with no changes except for the xref_tag
                  mentioned about (since some entries referenced other
                  entries).

* Once you click on sumbit, you will see the entry as it will appear to
  users, along with a submenu to Modify/Delete/Add entries. You can modify
  the current entry from that menu. Mere users do not see this menu, only
  when in reddot mode.

* The intent here is that we can generate new entries really easy, right
  from email if you like (with appropriate <pre> or <xmp> tags around it).

* I have added sql/knowlbase-create.sql and a makefile target to
  generate that file when creating a distribution. I also added a section
  to install/boss-install to insert the entries into the new DB.

* I hooked the search function into the existing Search Documentation link.
  We know search both the Knowledge Base *and* the Documentation on doc
  searches. This probably needs a little more work to get right.

* I changed a lot of faq links to be more consistent and to reference
  the proper xref_tags (#swapping instead of #UTT-34).

It breaks the windows port and also doesn't accomplish anything.
(a non-root user can always just make the tmcc call to boss to
get the info).

 * ipodinfo and watchdoginfo are now fullconfig commands
 * ipod setup moved later in startup to take advantage of fullconfig
 * watchdog changed to use cached info on the first call
 * move firewall setup later to take advantage of fullconfig
Other stuff:
 * accounts/ipodinfo cache files protected 600

Along with Tim's changes, this should eliminate 4 more boot time calls!

active users and projects lists; it was taking 3 seconds to generate
that table from the testbed_stats table. Switch to using newly added
last_activity slots of user_stats and project_stats tables. Now takes
less then .1 seconds ...

where we insert the testbed_stats record. This will make it much much faster
to generate the recent users list in genelists (currently taking 3 seconds!).

comment to more closely reflect what all is going on.