Closed kamatsuoka closed 11 years ago
+1
@kamatsuoka
Hi. Thanks for the pull request!
Excuse me, but according to your code and explanation, I guess that you can just use Elasticache with:
memcached.host="mycachename.asdfjk.cfg.use1.cache.amazonaws.com:11211"
Instead of which you are proposing:
elasticache.config.endpoint="mycachename.asdfjk.cfg.use1.cache.amazonaws.com:11211".
Are you just providing an alias(elasticache.config.endpoint) for the configuration key memcached.host
, or am I missing something?
I wanted to distinguish the config endpoint, which supplies a list of memcached hosts, from the memcached hosts themselves. Elasticache spares you the need to put the memcached hosts in your config, and it supports auto-reconfiguration of the cluster.
Also, in the memcached.host code path there are options for setting username and password for memcached. Those aren't supported in Amazon's memcached client.
I forgot to mention that you need to use Amazon's elasticache client instead of spymemcached if you want to take advantage of auto-discovery. So download AmazonElastiCacheClusterClient-1.0.jar (or latest version) from the AWS console, and remove the sbt dependency on spymemcached.
@kamatsuoka
I got it and I will publish a new version of the plugin with your patch, as soon as I have myself tried Elasticache a bit.
Thanks!
Takes advantage of Elasticache's auto-config mechanism: saves you from having to maintain the list of cache nodes in the client. To use, add your elasticache configuration endpoint to your conf as elasticache.config.endpoint={config.endpoint}. For example, elasticache.config.endpoint="mycachename.asdfjk.cfg.use1.cache.amazonaws.com:11211".