I am attempting to start up a 3-node cluster on one machine (Mac OS X Mountain Lion) for Riak via http://wiki.basho.com/Basic-Cluster-Setup.html
I downloaded and expanded the tarball for Riak 1.2 (no source to build, nice and clean expansion). I then was able to get a single node running no problem, with no special configuration. My Erlang version is R15B05, if you’re curious, but I don’t think this has anything to do with it, as 1.2 is fully supportive of this version.
I then recursively copied the directory riak-1.2.0 (the expanded tarball file) three times to create top level directories riak1, riak2, and riak3. I then proceeded to make the suggested changes from the link, changing the port numbers to be unique for each instance. However, when I changed the vm.args part (the name of the node from riak@127.0.0.1 to riak1@127.0.0.1 and the like) from the link, I received the following error in the error.log file:
2012-08-09 01:21:04.288 [error] <0.161.0> gen_server riak_core_capability terminated with reason: no function clause matching orddict:fetch(‘riak1@127.0.0.1’, [{‘riak@127.0.0.1’,[{{riak_core,staged_joins},[true,false]},{{riak_core,vnode_routing},[proxy,legacy]},…]}]) line 72
When renaming the node in vm.args (the -name param) back to riak instead of riak[new_number], everything works fine. However, as specified in the link, the nodes need to have different names to cooperate with one another. As shown in the error log, it seems like there is a name mismatch with something else when I customize the name? Is there some other config parameter that I haven’t seen that must be changed? I tried “raw_name” in app.config, but this didn’t make any difference. I appreciate any help I can get with this one. Thanks!
I had to do the following:
./riak1/bin/riak-admin reip riak@127.0.0.1 riak1@127.0.0.1
./riak2/bin/riak-admin reip riak@127.0.0.1 riak2@127.0.0.1
./riak3/bin/riak-admin reip riak@127.0.0.1 riak3@127.0.0.1
./riak4/bin/riak-admin reip riak@127.0.0.1 riak4@127.0.0.1