Hi, Ken Problem #1 is almost surely a hardware problem. Have you put ferrite coils around the cable from your logger to the computer? A UPS also helps. If you are still getting random corruption in the logger memory, then the logger may be defective.
Problem #2: a graphic is regenerated only as often as its aggregation interval. For example, if a month plot uses a 3 hour aggregation interval, then it will be regenerated only after it is 3 hours old. To force regeneration, delete the graphic. -tk On Thu, Apr 9, 2020 at 11:05 AM Ken Waters <[email protected]> wrote: > Hello, > > Sorry if this seems long but I tried to put everything in here that might > help explain the situation. > > Here is my setup: > Weewx version: 3.9.2 > Weather station: Davis Vantage Pro 2 connected thru DB9 serial port > Hardware: HP Pro 6000 SFF dual Pentium RAM: 4 GB, WD 1 TB drive > (currently only 2% usage) > O/S: Ubuntu 18.04.4 LTS > Website: http://satwatcher.us/satwatcher.us/wxstation/ > > I've been running WeeWx for about 9-10 years and previously used WxView. > > Here is the problem: > > Periodically and randomly the SQLite database gets corrupted with the most > recent update showing a future day/time. There seems to be no obvious > trigger that causes this and it's been happening to me for several years > including with other hardware setups. When this happens it basically > "locks" up the database so that current observations can no longer be > stored and results in a loss of data after the database corruption. In the > past I have had to resort to: > > a) stopping the service > b) clearing the memory (sudo /usr/share/weewx/wee_device -y > --clear-memory) > c) renaming the .sdb > d) restarting the service > > in effect, starting with a new database. This has been happening every > 2-4 weeks randomly and not as a result of anything done by me to the server. > > Here is a sample snippet from syslog that shows the issue: > Apr 9 07:38:15 WeewxWeatherServer weewx[28253]: vantage: Getting > archive packets since 2020-04-09 07:37:00 MST (1586443020) > Apr 9 07:38:15 WeewxWeatherServer weewx[28253]: vantage: Gentle wake up > of console successful > Apr 9 07:38:15 WeewxWeatherServer weewx[28253]: vantage: Retrieving 1 > page(s); starting index= 0 > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: manager: Added record > 2020-04-22 23:06:00 MST (1587621960) to database 'weewx.sdb' > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: manager: Added record > 2020-04-22 23:06:00 MST (1587621960) to daily summary in 'weewx.sdb' > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: *vantage: DMPAFT > complete: page timestamp 2020-04-07 12:59:00 MST (1586289540) less than > final timestamp 2020-04-22 23:06:00 MST (1587621960)* > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: vantage: Catch up > complete. > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: vantage: Requesting 200 > LOOP packets. > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: reportengine: Running > reports for latest time in the database. > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: reportengine: Running > report 'SeasonsReport' > Apr 9 07:38:16 WeewxWeatherServer weewx[28253]: reportengine: Found > configuration file /etc/weewx/skins/Seasons/skin.conf for report > 'SeasonsReport' > > Here is output captured this morning from a sqlite3 select command that > shows the erroneous timestamp in the database: > > $ sqlite3 /var/lib/weewx/weewx.sdb "select > datetime(dateTime,'unixepoch','localtime') from archive where > dateTime>1550000;" > 2020-04-09 07:30:00 > 2020-04-09 07:31:00 > 2020-04-09 07:32:00 > 2020-04-09 07:33:00 > 2020-04-09 07:34:00 > 2020-04-09 07:35:00 > 2020-04-09 07:36:00 > 2020-04-09 07:37:00 > *2020-04-22 23:06:00* <<<---- erroneous future entry > > These date/times are local MST (Z-7) and the problem actually occurred at > 2306 MST last night resulting in loss of data for about 8 hours. It seems > as if WeeWx is skipping recording the observation because it thinks this is > an old observation due to the last entry being "newer". > > Problem #2: > In addition, after stopping, clearing, removing the .sdb, and restarting I > now see that the graphics being produced are not current and are > time-stamped 4/22/2029 23:06 although the data on the left-side of the page > at http://satwatcher.us/satwatcher.us/wxstation/ are current. Why are > the graphics still keying off the erroneous time in the future? > > This has been driving me nuts for several years now. Does anybody have > any idea what may be causing this and how to fix it so it no longer happens? > > Thank you ever so much! > > (ps...I did post something to the list back in 2015 about this but it was > not resolved at the time) > > Ken > > > -- > You received this message because you are subscribed to the Google Groups > "weewx-user" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/weewx-user/2e5a14d8-a48d-47be-96b4-69625aa65d54%40googlegroups.com > <https://groups.google.com/d/msgid/weewx-user/2e5a14d8-a48d-47be-96b4-69625aa65d54%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "weewx-user" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/CAPq0zEBP8RSp-b5%3DQk%2BGsu9nkjg7GBcp%3Dfv5DN4Q3SR009k42g%40mail.gmail.com.
