Re: local truncate of hints

2013-11-14 Thread Cyril Scetbon
anyone ? -- Cyril SCETBON On 10 Nov 2013, at 10:48, Cyril Scetbon wrote: > Are you really sure of that ? cause Tracing says the opposite when I do it on > one node : > > cqlsh:system> TRUNCATE hints ; > > Tracing session: fe0dbc80-49dd-11e3-a39b-e367b713537d > > activity

Re: local truncate of hints

2013-11-14 Thread Boris Yen
I quick glimpse at the code. It seems that truncate only cares about if any node goes down. It does not check the replication strategy. Boris On Thu, Nov 14, 2013 at 9:30 PM, Cyril Scetbon wrote: > anyone ? > > -- > Cyril SCETBON > > On 10 Nov 2013, at 10:48, Cyril Scetbon wrote: > > > Are you

Re: local truncate of hints

2013-11-14 Thread Jonathan Ellis
You're right. You'd want to expose the node-local truncate over JMX. On Sun, Nov 10, 2013 at 3:48 AM, Cyril Scetbon wrote: > Are you really sure of that ? cause Tracing says the opposite when I do it on > one node : > > cqlsh:system> TRUNCATE hints ; > > Tracing session: fe0dbc80-49dd-11e3-a3