Hi Folks, We've noticed that when running Selenium Grid with Nutch we can easily kill an Amazon cluster. Amazon basically mention that load can get too high on the frontend with "...too many java, firefox, and screen processes" which stopped the EMR healthchecks from properly working. This caused the EMR "logic" to terminate the cluster. I wonder if anyone is or has undertaken JProfiling of Selenium + Nutch particularly on the Selenium Grid configuration. Thanks Lewis
-- *Lewis*
