quick follow-up -- of course using my own long-valued field seems to work 
just fine, but I would still like to understand what's going on here.  It 
might be nice to have the internally-generated behavior as a fallback?

On Sunday, April 6, 2014 1:18:55 PM UTC-4, Michael Sokolov wrote:
>
> I need to be able to provide timestamps that will be preserved by ES.  I'm 
> using the _timestamp field (although perhaps I shouldn't, because) it seems 
> that the values I provide are being ignored, and the current date-time is 
> being inserted instead. My mapping configuration looks like this:
>
> dynamic: strict
>
> _source: 
>   enabled: false
>
> _all: 
>   enabled: false
>
> _timestamp: 
>   type: date
>   enabled: true
>   index: not_analyzed
>   store: true
>
> properties:
>   ...
>
> I've tried providing the times as longs, as longs converted to strings, as 
> ISO-formatted date-times (YYYYMMDDThhmmss.SSS), but nothing seems to have 
> any effect -- I might as well provide 0.
>
> I'm inserting documents using the Java client API by creating a Map, with 
> key="_timestamp" and values as described above.
>
> Has anybody run across this before?  It seems like a very standard use 
> case if you want to keep a consistent timestamp across a cluster, or in any 
> way copy documents without updating their timestamps.
>
> -Mike
>

-- 
You received this message because you are subscribed to the Google Groups 
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/elasticsearch/b948a29d-92af-45c4-9491-97e9b6495615%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to