Re: [DISCUSS] KIP-424: Allow suppression of intermediate events based on wall clock time

[jo...@newrelic.com, jo...@newrelic.com]

On 2019/03/11 21:39:18, John Roesler <jo...@confluent.io> wrote: 
> Regarding flushing on shutdown, can you elaborate of the motivation for
> doing so?

I'm not in any way attached to this idea if you think it's not any additional trouble. It wasn't obvious to me how you would store the "wall clock time last flushed" in the changelog. 

Re: [DISCUSS] KIP-424: Allow suppression of intermediate events based on wall clock time

[jo...@newrelic.com, jo...@newrelic.com]

I'd like to move this KIP along. To summarize where we've landed so far:

- Should suppress() flush all records (would be _before_ the time elapsed)?
- Or should it preserve buffered records and reload on restart? For
this case, should the record be flushed on reload (elapsed time is
unknown) or should we reset the timer to zero?

John Roesler suggested we preserve the buffered records since suppression is backed with a durable store via the changelog. Works for me.

- How strict the guarantee when records should be flushed should be.

John Roesler suggested to use the system time we already checked at the start of the processing loop. This should incur no extra performance hits with extra calls to System.currentTimeMillis(). Works for me.

I'm happy to add these details to the KIP if all agree. Also encourage other questions before putting this to a vote.

Thanks!