If your client program does not use Auto Discovery, it can manually connect to each of the Memcached cache nodes. This is the default behavior for Memcached clients.
You can obtain a list of cache node hostnames and port numbers from the AWS Management Consoleaws elasticache describe-cache-clusters
command with the
--show-cache-node-info
parameter.
Example
The following Java code snippet shows how to connect to all of the nodes in a four-node cache cluster:
...
ArrayList<String> cacheNodes = new ArrayList<String>(
Arrays.asList(
"mycachecluster.fnjyzo.0001.use1.cache.amazonaws.com:11211",
"mycachecluster.fnjyzo.0002.use1.cache.amazonaws.com:11211",
"mycachecluster.fnjyzo.0003.use1.cache.amazonaws.com:11211",
"mycachecluster.fnjyzo.0004.use1.cache.amazonaws.com:11211"));
MemcachedClient cache = new MemcachedClient(AddrUtil.getAddresses(cacheNodes));
...
Important
If you scale up or scale down your cache cluster by adding or removing nodes, you will need to update the list of nodes in the client code.