GitLab

Fixed error in GatherSwapStats
(kludge-enum-function was being accessed as if it were
 a local scalar.)

(hopefully) fixed bug in startexp
affecting batch system by switching two cleanup checks.
(I wasn't giving it an opportunity to bail on a batchexp failing
 soon enough.)

Also, tidied up return code for tbswap.

files.

Don't reboot nodes when recovering from a failed assign during update.

Merge in my 'caveat' text

Remove descriptions of a couple dirs we aren't going to export

Update date in file

Mask off '64' bit from return from assign_wrapper, so
batch_exp doesn't choke on it.

Various Other changes to get Expt Modify ready for prime time.

 - If assign fails on a modify, experiment will
   be restored to old state, *not* swapped out.

 - Reboot option has been improved to reboot all
   nodes as part of os_setup, not in separate
   step.

 - Different assign error codes result in different
   retry behavior for assign_wrapper
   (Follow's Rob's change to assign to make it
    pass back special code for non-retriable faults)

 - '64' bit in assign_wrapper exit code indicates to tbswap
   that db/phys state hadn't been mucked with before
   the exit occurred
   (ergo, '65' and '1' are the common return codes,
    though the old 4,8,16,32 are still there for assign failing.)

 - (tbswap still returns codes from assign wrapper)

 - Added 5 sec pause between assign attempts.

 - Cleaned up tbswap code.

 - Physical state backup/restore removed from tbprerun,
   put into swapexp.

 - Interfaces table now getting cleaned up correctly
   (Mike noticed problem)

 - Changed menu display in showexp to show
   the "modify" menu option for swapped out experiments
   (like it used to.)

 - A couple other changes.

Note:
 Still admin-only, but I plan to change that soon.

To do:
 - Erase expt backups in /tmp after using them.
 - Re-viz failed experiments.

the idx to the corresponding experiment_stats record so you can see
the details for an experiment. Still needs a bunch of work.

from idle and swapped (but leave in active for now, until we move some
people out).

Sorry, but emulab-ops and testbed are hardwired strings.

already done.

Yeah, we should get around to adding FS-aware compression to
imagezip one of these days...

and a couple disclaimers about being out of date.

Added "Quick Tour" and "Source Tree Organization" to README.

doc directory.

a "source tree map" of sorts. Did some reformatting to overview, and added
a small para about idle detection to the expt life cycle para.

get an updated copyrights message.

but forgot to put in the real file.

any more!

Should run from cron, about every 15 minutes or so.

Sends mail to idle expts that haven't gotten mail lately. Grabs the magic
constants from the sitevars in the db, and respects idle_ignore. The
messages now get bcc'd to tb-automail instead of tb-ops, too.

With the current settings, it will send each idle expt a message every
four hours (like I've been doing manually, but more punctually and without
going away at night and on the weekend.)

It also has a mode for selecting just one pid/eid, and possibly forcing a
message to get sent. This part will replace the current swap request stuff
in the web (ie, this will be the backend instead of the web page sending
email directly).