Assuming your storage engine is bitcask: 1. Turn off all your Riak nodes 2. rm -rf --verbose /var/lib/riak/bitcask/*/* 3. rm -rf --verbose /var/lib/riak/anti_entropy/*/* 4. Turn Riak back on
This will preserve your buckets and bucket types in in cluster metadata. You can also automate that, but it's a little more complicated. On Sun, Jan 4, 2015 at 4:25 AM, Cosmin Marginean <cosmin...@gmail.com> wrote: > Hi, > > I’m running Riak 2.0.2 on Centos 6.5 installed from an RPM > (http://s3.amazonaws.com/downloads.basho.com/riak/2.0/2.0.2/rhel/6/riak-2.0.2-1.el6.x86_64.rpm) > > I’m interested in wiping the entire storage (in order to re-run tests, etc). > I have a script that traverses buckets/keys but I’m aware that this process > is costly, plus I might be dealing with some eventual-consistency issues. > > Consequently, I would prefer to wipe the “data” directory entirely if that’s > in any way possible, and re-start Riak. However, the structure seems to be > scattered in several directories and is not entirely clear which one needs > to be wiped for a “clean data reset” without corrupting anything: > > # ll /var/lib/riak/ > total 28 > drwxrwxr-x. 66 riak riak 4096 Jan 3 17:29 anti_entropy > drwxrwxr-x. 66 riak riak 4096 Jan 3 17:29 bitcask > drwxrwxr-x. 3 riak riak 4096 Jan 3 17:28 cluster_meta > drwxrwxr-x. 2 riak riak 4096 Jan 3 20:07 generated.configs > drwxrwxr-x. 2 riak riak 4096 Jan 3 17:29 kv_vnode > drwxrwxr-x. 2 riak riak 4096 Jan 4 07:23 riak_kv_exchange_fsm > drwxr-xr-x. 2 riak riak 4096 Jan 3 20:07 ring > > Thank you > Cos > > _______________________________________________ > 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
