Description of problem: When testing against a well loaded sam-1.1 system it is possible to get the thin servers to fail and crash. ~2000 systems > 60 orgs, each with no less than 4 environments > 15 roles > 50 users 4G of RAM 2xVCPU Version-Release number of selected component (if applicable): katello-cli-common-0.3.5-2.el6_2.noarch katello-headpin-all-0.2.13-1.el6_2.noarch katello-candlepin-cert-key-pair-1.0-1.noarch katello-glue-candlepin-0.3.3-1.el6_2.noarch katello-cli-headpin-0.2.2-1.el6_2.noarch katello-configure-0.3.7-1.el6_2.noarch katello-headpin-0.2.13-1.el6_2.noarch katello-selinux-0.2.4-1.el6_2.noarch katello-common-0.3.3-1.el6_2.noarch katello-certs-tools-1.1.5-1.el6_2.noarch + Performance Hotfix Steps to Reproduce: Against a loaded system (well loaded) run multiple, varying types of tests; best to use automation to drive the tests to ensure differing interactions with the environment. Actual results: thin server will crash without warning, HTTPS requests return Service Temporarily Unavailable. ps -ef|grep thin returns no pids service katello stop returns; /usr/lib/ruby/gems/1.8/gems/thin-1.2.11/lib/thin/daemonizing.rb:140:in `kill': No such process (Errno::ESRCH) Expected results: System should remain operational
Can't reproduce. Will be performance tested soon so if it is still a bug in 1.2 it should resurface