Pulse 66 Rover Data Observations
=============================================================================
2016-11-09 1800: Rover recovered

2016-11-09 1820: ~160 sites, 1750 images

2016-11-09 1840: Begin downloading data. WinSCP stops xfer syslog after 477MB

2016-11-09 1920: Switching to scp from Rover to Germanic with success.

2016-11-09 2000: After log_grinder, observed large heading error to starboard
                 beginning in late April and correcting itself in late Sept.

2016-11-09 2100: Leaving scp of entire deployment for overnight.

-------------------------------- 2016-11-10 ---------------------------------

2016-11-10 0620: Observed scp completed. Make backup copy of raw dataset.

2016-11-10 0650: Begin processing of dataset on Germanic.

2016-11-10 0740: Processing complete.

2016-11-10 0800: Move Rover from well deck to main deck. Observed left-most
                 valve on port chamber stuck open.

2016-11-10 0830: Copy images to flash drive for McGill

2016-11-10 0830: Split acm_calc.csv for loading into Libre.

2016-11-10 0850: Appears that around Apr 30, vehicle began veering sharply to
                 stbd during transits. Simple heading behavior does not
                 correct the issue.

2016-11-10 0920: Rover corrects veering issue around Oct 20.

2016-11-10 0945: Reviewing optode data. Observed during first measurement on
                 Nov 8 2016 port optode data matched port reference data.
                 Sampling monthly measurements. Many port optode dataset also
                 look compromised but not all.

2016-11-10 0955: Briefed Ken and Paul.

2016-11-10 1015: SW TODO - send optode data, shutdown, status report, modem
                 wakeup timing, test.
                 HW TODO - Port chamber valves, burn wires.

2016-11-10 1020: No random modem wakeup events logged in executive.csv, so we
                 can safely extend the modem wakeup timeout.

2016-11-10 1050: Used "cp -r --parents" embedded in a "find" command to
                 cp all the optode data into one folder. Tarred/gzipped and
                 sent to our shore contact girl.

2016-11-10 1100: SW TODO - shutdown, status report, modem wakeup time, test.

2016-11-10 1210: Alana produced optode data summary. Sweet.

2016-11-10 1400: poweroff, halt, reboot.

2016-11-10 1430: Modified executive.rb to trap SIGTERM and exit. Modified
                 supervisor.rb to set_timeout to 13 and `poweroff`.

2016-11-10 1500: Ran hibernation tests, 13 seconds is sufficient but does
                 not waste time.

2016-11-10 1530: SW TODO - status report, modem wakeup time, test.

2016-11-10 1630: Tested status report using remote_client. When only the
                 executive is running status does not include site number
                 nor iteration number. But when the supervisor is up and
                 running, both of those are included in the status report.
                 I think we just have to be patient. I will look into
                 somehow caching a status string. But I don't want to get
                 too clever here.

2016-11-10 1700: SW TODO - modem wakeup time, test.

2016-11-10 1830: Since there were no accidental modem wakeups during the
                 deployment, Paul suggests increasing the timeout time to
                 give us more time before the Rover goes back to sleep. I
                 chaged it from 170 seconds to 300.

2016-11-10 1830: SW TODO - status report, test.

2016-11-10 1930: Looking into simply storing the deployment status every
                 time it changes. Then the executive could simply read it
                 from the db.

2016-11-10 2100: Better results using the previous method. Noticed that the
                 executive reports the supervisor exits abnormally when it
                 is actually still running. Next status request is succeeds.

-------------------------------- 2016-11-11 ---------------------------------

0730: Closley watched traces on executive, supervisor, supervisor_server, and
      syslog during a remote_client conversation. Fixed "iteration-1 of 2"
      issue (remove spaces). Other than that, everything looked really good.
      Did see one supervisor exit, but it was another phantom.

0910: Tested modem comms with MF modem, deck box, MF dunker: Works without
      fail. Saved serial convo on Razorback\Documents.

1030: Tested modem comms with LF modem. The channel works fine, but the modem
      batteries are shot. Not enough juice to push through a full sentence.

1200: Prepared pulse 67 mission script from previous file. Full mission with
      launch and a benthic only version. Created two sleep-only scripts for
      use when we recover it in March. These missions can be started after
      homing the chambers to save energy on the way up and during recovery.
      All have been xfered to the vehicle and are ready to use.