Rain Rate record 15358.8 mm/hr on 14 May 05:48

[Glenn F]

It has probably been asked, and if so point me in the direction please.    I'd love to delete this record, I never built an ark nor got flooded out, so would love this to show the actual , real, value.. 

Running a FineOffsetUSB system .  

Glenn.. 

[Thomas Keffer]

Hello, Glenn

See the Wiki entry Clean up old "bad" data and see if that doesn't answer your question. If not, come back and we'll dig in deeper.

-tk

--
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 weewx-user+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Glenn F]

1. Cheers Thomas,....  if you have time to view the attched imge and see if I'm in the right direction,..  I got swamped in numbers,, love a little clarification... 
   

[Thomas Keffer]

It's hard to read, but that's probably right. You can confirm my just selecting the most relevant columns. Try this for your SELECT statement:

echo "SELECT dateTime, datetime(dateTime,'unixepoch','localtime'), rainRate from archive where rainRate > 100;" | sqlite3 /var/tmp/datbasebu/weewx.sdb

This print out just timestamps, the local time, and rain rates. Easier to read.

-tk

[Glenn F]

Ok that line of code was far better to establish the error, thanking you. ,,  I have performed the guide from " Clean Up Old Bad Data" and finger crossed that when it recycles the update it will be repaired,,  Toes crossed too.... 

[Glenn F]

OK   It would appear I have somehow mucked it up..  Station hadnt updated in 4 days so did a wee_device --clear which regenerated a current time and values but still does not update anymore and the offending record is still there..     At a loss.....

[Glenn F]

Attached the syslog which just looks error riden and a bit beyond me currently.. 

[gjr80]

Exactly what command have you used to start/restart WeeWX?

Gary

[Glenn F]

using the " sudo /etc/init.d/weewx start/stop "  command. I had this set perfectly bugger it until I tried to remove the bad data :( which is still there and now my webpage is only logging updates every half hr even with the same weewx.conf file, before it would give me the exact time update with a page refresh per minute , and also the plots are dots rather than the line graph it was before..  Almost feel like giving up on this years data , only had it since JAN, and just starting over.. Would be hair pulling if I had some :P 

On Sat, Sep 29, 2018 at 8:34 PM gjr80 <gjrod...@gmail.com> wrote:

Exactly what command have you used to start/restart WeeWX?

Gary

--
You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/JWXQzbrLbms/unsubscribe.
To unsubscribe from this group and all its topics, send an email to weewx-user+...@googlegroups.com.

[gjr80]

Ok, it looks like when you did the wee_device --clear you reset the archive interval on your station to the default 30 minutes. When WeeWX starts the archive interval is read from weewx.conf and the station (if it supports this), if there is a discrepancy then the station archive interval is used. In your case the answer is in the log (sorry did not read this far when you first posted, that was a lot of log!):

Sep 29 19:15:29 Weather-Station weewx[2068]: engine: The archive interval in the configuration file (300) does not match the station hardware interval (1800).
Sep 29 19:15:29 Weather-Station weewx[2068]: engine: Using archive interval of 1800 seconds (specified by hardware)=

This will explain the dotty plots as well, there is only one data point being saved every 30 minutes. I suspect you want to change to a 5 minute archive interval, to do this you will need to use wee_device again but this time with the --set-interval action, something like:

$ wee_device --set-interval=5

Note the interval is in minutes and also not you need to stop WeeWX before doing this and then start again once set. Depending on your privileges you may need to use sudo

Gary

[Glenn F]

Thanks Gary.   Website is back to running as it was. Though i used    "wee_device --set-interval=1" I need that immediacy. Happy.. Though I still have that fecking 15000mm/h  rain rate  ??   Clues../?

[gjr80]

Yep, what we need to do is find out where the bad data is. It will either be in the archive or in the daily summaries or both, judging by your site my guess is it's definitely in the daily summaries and maybe the archive. Tom gave you a command to run at post #4 above, that will check the archive. Try running it again, note you will need to stop WeeWX and make another copy of your database to /var/tmp (you can start WeeWX again as soon as you have made the copy):

$ echo "SELECT dateTime, datetime(dateTime,'unixepoch','localtime'), rainRate from archive where rainRate > 100;" | sqlite3 /var/tmp/weewx.sdb

What does it come up with, if that shows the bad data then we need to fix that. If it doesn't show the bad data then the issue is confined to the daily summaries. Let us know how you go.

Gary

[Andrew Milner]

generating an archive record every minute seems way over the top, and may create other issues if the reports cannot be generated within a 1 minute window - especially the reports generated once a day at midnight - as the database grows bigger and bigger (which will be very quickly at 1 minute archive intervals)

[Glenn F]

it provides an exact moment when calling up the website..   I have had zero trouble with that,,   I'd prefer to get a now factor than a 5min ago..  It works so , I will stay with it.