Hello again :) On Mon, Aug 27, 2012 at 4:45 AM, Yousuf Fauzan <yousuffau...@gmail.com> wrote: > Hello, > > My riak setup is on EC2 with EBS volumes storing data and ring state. This > ensures that I can teardown the cluster and restart it with all the data > intact. When I restart the cluster, I do a reip of all nodes on all nodes. > > This used to work with Riak 1.1.4 but ever since I shifted to 1.2, > riak-admin member_status shows all the nodes except one to be in legacy > mode. ring_status reports that nodes are "Currently in legacy gossip mode" > > Any ideas?
1.2 introduced overhauled cluster administration [1] via "riak-admin cluster" [2]. In this specific scenario you'll want to "join" the renamed nodes and "force-replace" the old nodes with the new ones (though I suspect that you've moved on to more pressing matters at this point). On a general note, while we didn't deprecate reip, try to avoid using it. The new commands are much more elegant. As far as "legacy" goes, that looks like something related to running pre 1.0 Riak... Hope that helps. Mark [1] https://github.com/basho/riak/blob/master/RELEASE-NOTES.md#overhauled-cluster-adminstration [2] http://wiki.basho.com/Command-Line-Tools---riak-admin.html#cluster > > -- > Yousuf > http://fauzism.co, > > _______________________________________________ > riak-users mailing list > riak-users@lists.basho.com > http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com > _______________________________________________ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
