On 9 February 2017 at 14:46, Daniel Miller <[email protected]> wrote:
> Hi Magnus, > > Thanks, this is great news! I was worried that clearing a planned node > removal would stop the node from which I was issuing the clear command, > which would be bad. Sounds like I have nothing to worry about in that case. > The behaviour of `riak-admin cluster plan` is quite complex, depending on the staged changes. If a `leave` operation has been staged, it will just undo the staged change; no node will be stopped. However, if a `join` operation has been staged, the joining node will be shut down after its ring has been cleared. When this node restarts, it will behave like a fresh unjoined node and can be joined again. If `riak-admin cluster clear` was run from a node that remains in the cluster, this node will be unaffected. > On running cluster plan multiple times to generate a new plan: this is a > little surprising since I think I’ve done that before but I haven’t > observed it generating a new plan. I’ll try it again. > You may see the effect more pronounced if you use `claim_v3`, which is more aggressive. For a discussion about how to rebalance the ring without adding or removing nodes, please see http://lists.basho.com/pipermail/riak-users_lists.basho.com/2016-November/018815.html > Thanks for very quick response. > Daniel > > Kind Regards, Magnus -- Magnus Kessler Client Services Engineer Basho Technologies Limited Registered Office - 8 Lincoln’s Inn Fields London WC2A 3BP Reg 07970431
_______________________________________________ riak-users mailing list [email protected] http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
