you can get away with loading from only one node if you're positive all
data is consistent. A repair prior to loading should be enough, but if that
doesn't work just load from all nodes.
On 11 Nov. 2017 23:15, "Brice Figureau"
wrote:
> On 10/11/17 21:18, kurt greaves wrote:
> > If everything goe
On 10/11/17 21:18, kurt greaves wrote:
> If everything goes smoothly the next incremental should cut it, but a
> full repair post load is probably a good idea anyway. Make sure you
> sstableload every sstable from every node if you want to keep consistency.
If the previous cluster had 3 nodes with
Hello
Thank you for your responses.
The question is still open: how can we debug this kind of problems and how
can we find what is cassandra doing/loading?
I don't mind to wait for loading data structures, but it will be great to
know what cassandra is loading and for how much time.
Thank you.
