Rebalancing

[coder]

If I understand the doc correctly, dynamic rebalancing feature when
adding new hardware is in progress. Until that feature, could I do the
following?

I start with a production setup consisting of single machine, say on
EC2.

If I want to increase my cluster size later, later I create more
machines on EC2 form a new cluster.

I run a batch program that will read all data from cluster 1 and
insert it into cluster 2.

I re-start app server nodes pointing to new cluster.

There will be brief outage but this is not something that will happen
all the time, so it's still manageable.

[Jay Kreps]

Yes, that would work. You can use the iterator interface on StorageEngine.

-Jay

[coder]

Thanks for that info... One more related question; What is the status
of the failover functionality? If one server in the cluster fails and
the replication factor is set to, say 2, will the cluster work without
data loss?

[Jay Kreps]

Yes, that works. Naturally the required-reads/required-writes must be
less than or equal to the number of remaining servers.

-Jay

[coder]

Awesome.. I have to say, I can't wait to use this stuff in production.
It's very clean and focused solution. My hat's off to you guys.

[jay....@gmail.com]

Thanks! Let us know if you run into any problems or have any questions
we can help with.

-Jay