Hacker News Reader

I don't know about "definitely." Swap death is often worse than a reboot because (a) response latency goes through the roof, effectively making the service down anyway, and (b) you probably can't log in anyway because of I/O saturation (ssh, shell etc. could well be swapped out).

View on HN

(c) a load balancer can often understand a server that's down, a server that might end up doing I/O while holding locks probably wouldn't be detected.

Indeed. I've had zLinux boxes using VDISKs for swap that get it to the terrible state where the VDISK can sustain swapping at a high enough rate that Oracle will be able to accept and maintain connections, but not do any work. The application attached to the database is unable to see this as a failure and flop over to the other node in the RAC.

Sounds like underspecified failure conditions. Rarely are they as clean as, "I am dying, poof!"

Reply on news.ycombinator.com