>From weewx user guide: However, if the internal memory gets garbled, the console will immediately return archives in the past, and so it looks like the timestamps have decreased in value and so weeWX figures that is it: there is no more data.
I believe this is what happened. Previously my weewx would start up on powerup with times in 2016 and corrupt the data logger memory. Now I have modified it to start only if time is in 2018. But I believe weewx stopped filling missing archives as it found decreasing timestamps. Clearing my logger memory (anyway it has only 2 weeks worth contents) should fix this issue for good I guess. The catchup messages come in log only if wee_debug is set to 1 I believe? Anyway will look at the log again tonight. What is still a mystery is why did the pi go down and why did it need a power on cycle to start up again! On Monday, 29 January 2018 16:10:23 UTC+5:30, Praveen Chandrasekaran wrote: > > One additional detail - my pi is read only mode. On abrupt power down the > latest SQL database may not have been stored. Not sure if that messed up > anything though I believe the pi should have still rebuilt the records. > > Right now I am using ntp with fake hw clock removed, I am planning to put > in a RTC soon. > > Will dig more in the log sometime and also post the log here. > > On Monday, 29 January 2018 15:01:56 UTC+5:30, Andrew Milner wrote: >> >> The log would have shed light on any problems. >> You should not have needed to use any utility at all - what made you >> decide to use wee_database??. >> If you had done nothing weewx would have retrieved records from the >> logger from after the last archive record in its database automatically - >> this sometimes takes some time. On a RPi you should either add a RTC >> module or completely disable the fake hw clock to ensure weewx does not >> start until it has got a correct time from the NTP servers to ensure >> correct automatic operations which are reliant on accurate timestamps (such >> as recovery from a power outage). >> >> The log is always the key - and without a log everything becomes >> guesswork and speculation whilst with the log we can be talking actual >> facts!!!! >> >> >> >> On Monday, 29 January 2018 09:57:59 UTC+2, Praveen Chandrasekaran wrote: >> >>> Thanks. Will post with all details from next time. For now I did a >>> wee_device --dump and that populated the sql database after which I could >>> update WU with wunderfixer (the summary still has not got updated in WU as >>> per updated records, I can see updated records in graphs and table - I >>> believe that is an issue at WU end and I have filed a bug report with them >>> for the same). Not sure why weewx did not populate the sql on its own >>> though without forcing through wee_device --dump. >>> >>> On Sunday, 28 January 2018 23:22:40 UTC+5:30, Tom Keffer wrote: >>>> >>>> This should happen automatically, but it's hard to say what the problem >>>> is without a log. See *Help! Posting to weeWX >>>> <https://github.com/weewx/weewx/wiki/Help!-Posting-to-weewx-user>*. >>>> >>>> -tk >>>> >>>> On Sun, Jan 28, 2018 at 10:15 AM, Praveen Chandrasekaran < >>>> [email protected]> wrote: >>>> >>>>> Hi, >>>>> >>>>> For some reason my pi went down yesterday morning for 5 hours. I tried >>>>> rebuilding database using wee_database but to no avail. How do I import >>>>> data from data logger to weewx so that I can use wunderfixer then to >>>>> publish again to wu? >>>>> >>>>> Regards, >>>>> Praveen >>>>> >>>>> -- >>>>> 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. >>>>> >>>> >>>> -- 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.
