I'm running my cassandra cluster (one node), it went fine for a while under heavy load , using 270% CPU etc (this is a multi-core machine). but after a while, the CPU utilization dropped to 100%, and the cluster stopped responding to requests.
I did a jstack , and did ps ax -L -o tid,pid,%cpu,command|less I found that one thread is taking 45% of CPU, and the corresponding entry in the jstack output is: "VM Thread" prio=10 tid=0x09a50400 nid=0x1a1a runnable what is the "VM Thread" ? youkit dump does not show a thread with this name Thanks Yang
