Gary, I'm all set! Thanks for the advice on doing the dump first. I did that and it loaded all of the archive data into the database - as you said, dump generated tons of "duplicate record" errors in the log. After the dump, I then tried unplugging and pulling the batteries out of the console. After starting weewx again, the log showed errors reading LOOP data, so I stopped weewx again and ran wee_device --clear-memory as instructed in the link you sent. I restarted weewx and all is better. WooHoo! Doesn't look like I lost anything except app_Temp (we just discussed that on the new Belchertown topic thread!) over that period. Since weewx calculates appTemp on the fly as it writes the archive values and it is not in the console memory, I wouldn't expect it to be written when a dump is used. It's not really a concern to have missed some of those values. I'm really interested in Apparent Temperature as a real time number, not so much its historical values.
Interestingly, the reason I didn't notice the console data hang-up for 30 hours is that I see realtime data on my web page fed from MQTT. That and the fact that I live in Western North Carolina in the US where we just had our winter storm of the decade that dropped more snow in a day than we normally get in a year! So, I wasn't looking much at the graphs. Thanks again. phil On Monday, December 10, 2018 at 3:56:07 PM UTC-5, gjr80 wrote: > > Hi, > > If it's a Davis and there has been a power outage chances are it is > corrupt station memory. Hopefully though you won't have lost 30 hours of > data (assuming you had batteries in the console), perhaps just one or two > records. You can sometimes get a pretty good indicator from the log if this > is the issue; typically the station will still emit loop packets but not > archive records, so if you are using hardware record generation there will > be no archive records being saved (so you won't see such entries in the > log) but anything based on loop packets will still be working; eg WU RF > updates or any services that are bound to new loop packets arriving. > > In any case the solution is the same, follow the steps, in order, for > corrupt station memory under WeeWX generates HTML pages, but it does not > update them > <http://weewx.com/docs/usersguide.htm#html_generated_but_not_updated>. > Note the comment about dumping station memory at the end of the section. > Make sure you do this first, this is what may well save you missing data. > If the logger memory can be accessed you will likely see a lot of duplicate > record errors in your log, that is fine and expected, but it should also > get the missing records and save them to database. > > Good luck. > > Gary > > On Tuesday, 11 December 2018 06:01:39 UTC+10, Philip Kutzenco wrote: >> >> I had a power outage about 30 hours ago and just realized that, since >> then, my weewx.sdb database is not being updated. I first thought that the >> database was damaged when the Raspberry Pi running weewx took the hit. >> However, I can open the database in sqlite3 and it looks ok. I also ran >> wee_database --reconfigure. I renamed the newly built database, but it's >> still not being updated by weewx. >> >> So, I'm assuming my Davis Vantage Vue console (with a USB data module >> connected to the Raspberry Pi) is what got hung. That console, while >> plugged in, does also have batteries, but I'm still guessing it took a hit >> from the power outage. I know that I can reset the memory in the >> console/data module, but I'd lose 30 hours of data (not really the end of >> the world). However, is there anything I can do to avoid losing that data? >> >> Thanks in advance. >> phil >> > -- 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]. For more options, visit https://groups.google.com/d/optout.
