Re: [weewx-user] Re: Davis Datalogger bug

vince writes: >- 'test' your system's powerout behavior. Every computer comes up at a >predictable date+time when it is powered on, if you do not have a RTC. >See what that value is. On BSD, the default behavior on poweron is more or less to use time from the root superblock, s

Re: [weewx-user] Re: Davis Datalogger bug

On Thursday, October 17, 2019 at 11:15:00 AM UTC-7, David VE3STI wrote: > I have used the RPi because it is a power miser, it is small and it is > cheap enough to be bought new and 'dedicated' to a single purpose. > Yup. Most pi users do the same, as do people with similar tiny/low-power non-

Re: [weewx-user] Re: Davis Datalogger bug

This is veering into the tangential but... My weather logging needs are modest - I'm usually looking for the weather 'at the moment' or for the last several days. Occasionally, rainfall for the month. My livelihood does not depend on my weather observations. If I miss weather data during a power o

Re: [weewx-user] Re: Davis Datalogger bug

On Thursday, October 17, 2019 at 5:41:32 AM UTC-7, Thomas Keffer wrote: > > Yes, that's the problem. > You cannot expect high-quality data without investing in the hardware. You > bought an expensive Davis station, why not spend the other $50 to protect > the data? > > > Or alternately sell all

Re: [weewx-user] Re: Davis Datalogger bug

Yes, that's the problem. You cannot expect high-quality data without investing in the hardware. You bought an expensive Davis station, why not spend the other $50 to protect the data? I have a Davis Envoy, a UPS, and a high-quality embedded PC with a clock. It has run continuously without a singl

Re: [weewx-user] Re: Davis Datalogger bug

I'm sorry, but I didn't understand in which way the clock is related to this problem Il giorno giovedì 17 ottobre 2019 13:59:09 UTC+2, Andrew Milner ha scritto: > > regardless I would recommend that all Rpis controlling weather stations > have an RTC installed to avoid temporal issues as much as

Re: [weewx-user] Re: Davis Datalogger bug

I will have to review that section of the User's Guide. I caved and began using a UPS quite some time ago - but I think it is going to need a new battery soon. At this time, I'm some 1,000 km from home so no fiddling with WeeWX for a while. When I'm home, I am going to try Thomas Keffer's suggesti

Re: [weewx-user] Re: Davis Datalogger bug

Il giorno giovedì 17 ottobre 2019 14:13:51 UTC+2, Thomas Keffer ha scritto: > > As it is suggested in other answers here, it would "fix" the problem >> buying an UPS, but I honestly wolud find a software solution, not coming >> around the problem in this way (which could also be expensive). >>

Re: [weewx-user] Re: Davis Datalogger bug

> > As it is suggested in other answers here, it would "fix" the problem > buying an UPS, but I honestly wolud find a software solution, not coming > around the problem in this way (which could also be expensive). > There is a software solution: run with record_generation = software. This essentia

Re: [weewx-user] Re: Davis Datalogger bug

regardless I would recommend that all Rpis controlling weather stations have an RTC installed to avoid temporal issues as much as possible. On Thursday, 17 October 2019 14:26:18 UTC+3, Andrea Cecilia wrote: > > I've never seen this section of the User's Guide (it might be relatively > new), bu

Re: [weewx-user] Re: Davis Datalogger bug

I've never seen this section of the User's Guide (it might be relatively new), but it is exactly about the problem we are discussing here. So, the fixing tries they suggest without data loss are two: 1) unplug and reboot the console by removing current and batteries for about 2 minutes and then

Re: [weewx-user] Re: Davis Datalogger bug

If you care about data integrity, the $50 for a modest UPS is money well spent! On Wed, Oct 16, 2019 at 5:00 PM gjr80 wrote: > Yes, the price you pay for using software record generation. Also, I > forgot about no_catchup being introduced in 3.9.1. That changes the logic > during a startup catch

Re: [weewx-user] Re: Davis Datalogger bug

Yes, the price you pay for using software record generation. Also, I forgot about no_catchup being introduced in 3.9.1. That changes the logic during a startup catchup, now software record generation will also get caught up in the 'timestamp less than final' issue unless you set no_catchup = Fal

Re: [weewx-user] Re: Davis Datalogger bug

Immune, yes, but you also won't get any archive records stored on the logger while your computer is down! On Wed, Oct 16, 2019 at 4:33 PM gjr80 wrote: > One other thing to keep in mind; the de facto standard mode of operation > of a Davis station under WeeWX is to use hardware record generation,

Re: [weewx-user] Re: Davis Datalogger bug

One other thing to keep in mind; the de facto standard mode of operation of a Davis station under WeeWX is to use hardware record generation, in other words WeeWX obtains and stores archive records from from the console/logger. WeeWX can also operate using software record generation where only

Re: [weewx-user] Re: Davis Datalogger bug

This has been my solution as well: a real-time clock, plus a UPS. But, if the power outage is severe enough to drain the UPS battery, my setup still recovers. At least, the last time that happened (about a year ago). -tk On Wed, Oct 16, 2019 at 8:14 AM David VE3STI wrote: > I had somewhat simi

Re: [weewx-user] Re: Davis Datalogger bug

I had somewhat similar issues - WeeWX couldn't read from the Vantage Vue after boot up. I needed to do the "memory clear" command then all was well. However, it was a giant pain if I was away/busy/etc and not available to run the command. I thought that it should be possible to "fix" this issue

Re: [weewx-user] Re: Davis Datalogger bug

This is likely a case of memory corruption, brought on by time not being synced during the reboot, due to the lack of an on-board clock. It was not a problem with Weatherlink because you were not running it on an RPi. See the section *WeeWX generated HTML pages, but does not update them

[weewx-user] Re: Davis Datalogger bug

what came just before the log you posted? why is weewx attempting to do a catchup? had weewx just been started? had communications been lost for some reason? We need to know what is causing weewx to attempt a catchup in the first place. You can never post too much log On Wednesday, 16

[weewx-user] Re: Davis Datalogger bug

here it is. As you can see, there is no "added record [...] to weewx.sdb" line, which indicates that it is not downloading data from the datalogger, and the reson (I think) is explained in the highlighted line. Oct 16 10:30:14 raspberrypi weewx-vp2[511]: vantage: Getting archive packets since 2

[weewx-user] Re: Davis Datalogger bug

attaching the log not just a line may let people offer more advice. On Wednesday, 16 October 2019 13:05:21 UTC+3, Andrea Cecilia wrote: > > Hi there, > I'm working with lot of Davis stations connected to weewx, and I noticed > that there exists a frequent bug which I'm trying to describe now: