G'day!
I'm currently trying to set up a test environment mimicking our
production environment by copying the leveldb directory from each
production node onto the corresponding test node. By carefully matching
the node's partition ownerships to determine where I copy the data I can
start up Riak on the test nodes and not have any partition handoffs and
it seems like I have a working snapshot of my production data.
The questions I have are:
1: Is copying the leveldb directory from one cluster node to a node in a
completely different cluster a sane thing to do? It seems to work. I
don't get any errors in the Riak or LevelDB logs and my data is there
when I query it.
2: Is there a simple way to determine which node owns which LevelDB
partition? Currently I just guess. It looks like partitions are
allocated to nodes alphabetically by node name. So I can copy the
leveldb directory from production node "apple" to test node "pear",
production "banana" to test "quince" and so on and it works. If I copy
"apple" to "quince" and "banana" to "pear" then I see partition handoffs
between "pear" and "quince" when I start Riak. It would be good to know,
empirically, that partition "1234...789" on "apple" belongs on "pear"
and not "quince".
3: Is there a better way to do this? "riak-admin reip" is deprecated.
"riak-admin force-replace" doesn't work because the production node
doesn't exist in the test cluster.
Thanks!
Shane.
_______________________________________________
riak-users mailing list
[email protected]
http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
