Our hadoop cluster rolled over last friday. As a result, we've got several procs in sql server 2014 who are querying the hadoop cluster via linked server. When i try to kill the spids who are using this linked server they get stuck in a roll back state. Anyone know how to kill these spids w/o bouncing the sql service? I was reading about DTC and DAC however we did not have DAC enabled. Ideas?
↧
