[codewright]

Your Solr cluster kicked the bucket at 150QPS?

Jesus dude. I couldn't reproduce that with my single or multi-node ElasticSearch clusters if I wanted to.

How were the EBS backing stores setup for these EC2 nodes?

Edit: Also, when I was talking about "them" falling apart, I meant Postgres or Sphinx, not Solr/ElasticSearch.

Well-configured Solr and ElasticSearch clusters can work very well for most people.

[fizx]

EBS shouldn't really matter, because with a reasonable heap, he should have 40-50G of available filesystem cache, and 48G of data.