A very unusual event happened during the telemetry of profile 5217.004 as shown in the following snippet from the engineering logs:

(May 01 2007 07:53:40,  882827 sec) TelemetryInit()      Profile 4. (Apf9i FwRev: 012407)
(May 01 2007 08:00:44,  883251 sec) AirSystem()          Battery [189cnt, 14.7V]  Current [82cnt, 33.0mA]  Barometer [125cnt, 6.2"Hg]  Run-Time [51s]
(May 01 2007 08:01:13,  883279 sec) GpsServices()        GPS almanac is current.
(May 01 2007 08:01:13,  883279 sec) GpsServices()        Initiating GPS fix acquisition.
(May 01 2007 08:01:29,  883295 sec) gga()                $GPGGA,080526,2013.4285,S,09556.6298,E,0,00,,,M,,M,,*48
(May 01 2007 08:01:39,  883305 sec) gga()                $GPGGA,080536,2013.4285,S,09556.6298,E,0,00,,,M,,M,,*49
(May 01 2007 08:01:49,  883315 sec) gga()                $GPGGA,080546,2013.4285,S,09556.6298,E,0,00,,,M,,M,,*4E
(May 01 2007 08:02:01,  883328 sec) Apf9Init()           Asynchronous wake-up detected (ie., wake-up not initiated by alarm signal).
(May 01 2007 08:02:08,  883335 sec) AirSystem()          Air-bladder inflation by-passed. Barometer: [126cnt, 6.5"Hg].
(May 01 2007 08:02:33,  883360 sec) Telemetry()          Unable to open "5217.004.msg" in append mode.
(May 01 2007 08:02:34,  883361 sec) GpsServices()        GPS almanac is current.

It appears that a spontaneous reset was generated during fix-acquisition (watch dog?). Afterwards, the hydrographic data file (5217.004.msg) couldn't be opened because a FILE* object in persistent RAM was still attached to the file. Fault tolerance measures that were built into the firmware during the initialization of each profile cycle eventually caused the 5217.004.msg to be closed without loss of data.

Hence, the firmware responded as intended to a file-system exception with potentially a broad-spectrum of causes. Profile cycle 005 executed correctly and without unusual exceptions.