Import is so fast you’re going to spend more time worrying about what ifs 
than just importing a week of data. If it’s really the same actual station 
then a db dump on system A and a db restore of the dumped data into systemB 
get you there very quickly.

On Monday, March 17, 2025 at 9:14:24 AM UTC-7 Ryan Stasel wrote:

> Thanks gjr80. I guess the question is, how do I tell easily what data 
> already exists (the time stamps will have to match, I assume). So 
> "2025-02-28 08:15:00" existing (for example) I assume won't match against 
> an import of "2025-02-28 08:16:00" (let alone if there's seconds other than 
> 00). 
>
> Is there an "easy" way to tell? or am I stuck going into sqlite3 and doing 
> some selects (not hard, just was kind of hoping the import process WOULD 
> report on records already existing). =)
>
> On Sunday, March 16, 2025 at 5:59:03 PM UTC-7 gjr80 wrote:
>
>> The default weectl import action is to *not* import records where there 
>> is already an archive record in the database with the same timestamp. weectl 
>> import will appear to import such records but if you look at the WeeWX 
>> log you will see entries similar to:
>>
>> 2023-11-04 15:33:01 weectl-import[3795]: ERROR weewx.manager: Unable to 
>> add record 2018-09-04 04:20:00 AEST (1535998800) to database 'weewx.sdb': 
>> UNIQUE 
>> constraint failed: archive.dateTime
>>
>> indicating a record with the same timestamp (in this case 2018-09-04 
>> 04:20:00 AEST (1535998800)) already exists in the database and the imported 
>> record was discarded. Long story, but it was too inefficient to check every 
>> record before attempting to add it to the archive, hence the somewhat 
>> cumbersome import and notification. Note that if you use the (presently) 
>> undocumented --update command line option with weectl import 
>> pre-existing records will be overwritten.
>>
>> Bottom line - just run weectl import as per the docs, and whatever you 
>> do don't include --update on your weectl import command line. And of 
>> course always make a backup of your database before importing.
>>
>> Gary
>> On Monday, 17 March 2025 at 09:56:17 UTC+10 [email protected] wrote:
>>
>>> I lost a few days of data due to a network issue, but I have the data in 
>>> another system. Am trying to import from csv, but I think my csv overlaps a 
>>> bit with what's already in the weewx database. 
>>>
>>> Doing a dryrun, it says it will import every line in the csv... 
>>>
>>> Is there a way to get weectl import to NOT import duplicate data, or is 
>>> there an easy way to see what data is in those dates so I can modify csv to 
>>> exclude those entries?
>>>
>>> Thanks! 
>>>
>>

-- 
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 visit 
https://groups.google.com/d/msgid/weewx-user/01a2b7b6-a617-4915-935e-c312d62ba204n%40googlegroups.com.

Reply via email to