[erlang-questions] Large scale deployments testing

[Michael Truog]

On 09/21/2012 03:00 PM, Joel Meyer wrote:
> Hi Sergey,
>
> We (OpenX) have a riak_core based application that's running on a 125 node cluster (there are also other smaller clusters). We never really tested to see where it would fall over (and the cluster was much smaller when it started), but I see no indicators that it will fall over when we add the 126th node. FWIW, it's running riak_core 0.13.0, and I assume the newer versions of riak_core have only gotten better. Answers to some of your other questions (based solely on my experience) in-line below.

The riak list mentioned optimism about riak clusters of 200+ nodes in a thread here:
http://riak-users.197444.n3.nabble.com/does-Riak-cluster-maintain-fully-connected-Erlang-network-td3695942.html

However, as mentioned in the thread, a fully connected network of nodes (fully connected because of the usage of distributed Erlang) does  have a natural limit (due to the network speed) on scalability with the net tick time.  You can always increase the net tick time, but then failures will take longer to detect.

So, your success may rely on your fault-tolerance requirements.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20120921/51affc9c/attachment.htm>