This morning we had a node failure in the web app tier which resulted in several minutes of "server not accessible" messages for our readers. I've switched in a replacement node and now I am performing an autopsy on the old node.
Current status: all systems go
Comments