Jeff, Benedict, we’ve taken your thoughts on this on board and will probably plan to repair in between scaling. The point about speculative retry did get me thinking. Many thanks for the tip and the explanation!
> On 23 Aug 2021, at 3:46 pm, Jeff Jirsa <jji...@gmail.com> wrote: > > > >> On Aug 22, 2021, at 7:25 PM, Miles Garnsey <miles.garn...@datastax.com> >> wrote: >> >> >>> >>> The problem is that today there’s no way to reliably exclude the new DC >>> from serving reads, that I know of? If you can, then yes you would only >>> need to ensure repair were run prior to activating reads from this DC. >> >> We think we have a way to do this using certain settings in the Java driver. >> >> Agree on your other points! > > I don’t see how > > Your best chance is with snitch games > > And those don’t guarantee correctness if a single replica GC pauses and > forces a speculative retry > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@cassandra.apache.org > For additional commands, e-mail: dev-h...@cassandra.apache.org >
