We are running a full Brisk ring on ec2 (all on m1.larges) each running JT and Cassandra. I had one machine hit 91% actual memory utilitzed last night setting off the pagers. My question is, where do we override the memory settings? Should they be set in cassandra-env.sh or do you do something special in brisk-env.sh?
Also, has anyone figured out what the proper settings are in our case? We only run brisk, a lightweight monitor, and puppet for config management on these machines (debian 6).