Change Details

In our meeting this morning we agreed that the cassandra / restbase cluster will initially be run by services. This means that both @mobrovac and @gwicke will need sufficient access to the cluster to: - restart cassandra - disable puppet & test changes in the cassandra config on one of the boxes - restart restbase - attach the node debugger to restbase, perform heap dumps - run some root-only debug tools like netstat -p The simplest way to support this (at least initially) is to give us local root on these boxes. Since we are going to be responsible for running this cluster, I think this is reasonable.

In our meeting this morning we agreed that the cassandra / restbase cluster will initially be run by services. This means that both @mobrovac and @gwicke will need sufficient access to the cluster to: - restart cassandra - temporarily disable puppet & test changes in the cassandra config on one of the boxes - restart restbase - attach the node debugger to restbase, perform heap dumps - run some root-only debug tools like netstat -p The simplest way to support this (at least initially) is to give us local root on these boxes. Since we are going to be responsible for running this cluster, I think this is reasonable.

In our meeting this morning we agreed that the cassandra / restbase cluster will initially be run by services. This means that both @mobrovac and @gwicke will need sufficient access to the cluster to: - restart cassandra - temporarily disable puppet & test changes in the cassandra config on one of the boxes - restart restbase - attach the node debugger to restbase, perform heap dumps - run some root-only debug tools like netstat -p The simplest way to support this (at least initially) is to give us local root on these boxes. Since we are going to be responsible for running this cluster, I think this is reasonable.