Entry | Date | Title | Site | Author | #Graphics |
---|---|---|---|---|---|
89 | Sat 29-Apr-2006 | Backup to USB disk | Manzanar home TREX | Bill | |
65 | Mon 10-Apr-2006 | KVH replaced, then disabled | Kiesarge Mine Site | Gary | |
58 | Fri 07-Apr-2006 | NTP crontab entry | MISS | Gary | |
55 | Fri 31-Mar-2006 | Radar parameter manual navigation | MISS | Gary | |
52 | Thu 30-Mar-2006 | EOP software checks (NIMA) | Manzanar home TREX | Gary | |
20 | Tue 14-Mar-2006 | data_store backup: 7pm Mar 13 (03UT Mar 14) | MISS | Steve |
Doing backup to an external UBS disk. Tried connecting using firewire (daisy chained to Maddog DVD drive) but it wasn't recognised. So disconnected printer from USB port and plugged in disk there and a Konqueror window popped up. Backing up (from /data_source directory) using command: rsync -v -u -r -t * /media/usbdisk/data/miss/data_source/
I replaced the KVH with the one that Lou shipped to us. After calibrating it at the Independence Airport, it still doesn't work. Then at this site the ingestor went berzerk repeatedly reporting "unexpected data" from the KVH, so I've just disabled kvh ingest in the configuration. The profiler antenna orientation can continue to be measured manually with the compass and then entered into the radar parameter file navigation script.
Since the satellite internet is never up before the laptop boots, the NTP startup does not synchronize with its servers. Restarting the ntpd service always fixes this, so I just added a crontab entry to restart ntpd every hour. That seemed harmless enough. Here's the crontab line: 30 * * * * /etc/init.d/ntpd restart > /dev/null 2>&1 It might be that ntp would eventually synchronize with the servers and I just wasn't patient enough to notice. An easy way to check is to run ntpq and the 'lpeers' procedure: [iss@iss1 tklog]$ ntpq ntpq> lpeers remote refid st t when poll reach delay offset jitter ============================================================================== ntp-1.cns.vt.ed 198.82.247.40 2 u 39 64 7 767.470 -13.153 46.794 ntp-2.cns.vt.ed 198.82.247.40 2 u 37 64 17 790.262 -9.692 26.999 ntp3.tamu.edu 128.194.254.7 2 u 36 64 7 882.431 -3.500 38.574 ntp1.tummy.com 132.163.4.101 2 u 34 64 7 784.581 -32.636 43.044 otc2.psu.edu 18.26.4.105 2 u 36 64 7 854.045 11.158 24.974 mapr.localdomai .STEP. 16 u - 64 0 0.000 0.000 4000.00 iss-mapr.locald .STEP. 16 u - 64 0 0.000 0.000 4000.00 ntpq>
The script which automatically sets the navigation settings in the POP radar parameter files now accpets manual overrides of the KVH trailer bearing and also an altitude. This makes the following procedure possible when the KVH is not working: Read the trailer bearing in degrees from true north (the direction in which the tongue trailer is pointing). Stop POP. Double-click the ruler icon on the left side of the desktop, titled Navigate Radar Parameters. This runs a script which prompts for the trailer bearing and the altitude. The most recent settings will be the defaults. The altitude (in meters) can be read from the GPS main screen. Then allow the script to modify all of the radar par files on the PC and replace the existing files. If the KVH is correct, the trailer bearing can be specified as 'auto'. If an altitude is not given (ie, it is set to less than zero, like -999), then the altitude in the par files is not changed, and the correct altitude must set explicitly with the POP Alt-F3 screen.
NIMA results have not been sent for 3/28 and 3/29, so I came to check on it. I noticed the gps ingestor (gps_adam) was taking up lots of cpu, 80-90%. strace showed it looping through select and read: select() was returning that the serial port was ready, but a subsequent read returned 0 bytes: select(6, [3 5], NULL, NULL, {300, 0}) = 1 (in [5], left {300, 0}) read(5, "", 512) = 0 select(6, [3 5], NULL, NULL, {300, 0}) = 1 (in [5], left {300, 0}) read(5, "", 512) = 0 ... Restarting the gps appears to have fixed it. Perhaps this a bug in the edgeport or serial driver. I then noticed 3 wppp processes running at the same time for the same day, so something got hung up or slowed them down. I killed all of those and restarted the processing for 3/29. According to the log for 3/28, a profiler parameter change is preventing that whole day from being processed. It took about an hour to process one whole day, so I increased NIMA run times to every four hours. There were old data files from 2/18 with zero length which were causing error log messages in zebra, so I removed them.
Charlie wrote the data_store to a DVD at 7pm Monday March 13 (03Z Mar 14). I will bring this DVD ack to Boulder