(1) We have three NodeManager. [ec2-user@ip-10-0-0-4 ~]$ yarn node -list -all 13/10/24 17:50:59 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable 13/10/24 17:51:01 INFO client.RMProxy: Connecting to ResourceManager at ip-10-0-0-10.us-west-2.compute.internal/10.0.0.10:8032 Total Nodes:3 Node-Id Node-State Node-Http-Address Number-of-Running-Containers ip-10-0-0-101.us-west-2.compute.internal:42205 RUNNING ip-10-0-0-101.us-west-2.compute.internal:8042 0 ip-10-0-0-103.us-west-2.compute.internal:60803 RUNNING ip-10-0-0-103.us-west-2.compute.internal:8042 0 ip-10-0-0-102.us-west-2.compute.internal:59201 RUNNING ip-10-0-0-102.us-west-2.compute.internal:8042 0 ===== (2) We rebooted NodeManager, and we can confirm information before the reboot as "LOST". [ec2-user@ip-10-0-0-4 ~]$ yarn node -list -all 13/10/24 18:20:23 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable 13/10/24 18:20:25 INFO client.RMProxy: Connecting to ResourceManager at ip-10-0-0-10.us-west-2.compute.internal/10.0.0.10:8032 Total Nodes:6 Node-Id Node-State Node-Http-Address Number-of-Running-Containers ip-10-0-0-101.us-west-2.compute.internal:48281 RUNNING ip-10-0-0-101.us-west-2.compute.internal:8042 0 ip-10-0-0-103.us-west-2.compute.internal:51291 RUNNING ip-10-0-0-103.us-west-2.compute.internal:8042 0 ip-10-0-0-102.us-west-2.compute.internal:52128 RUNNING ip-10-0-0-102.us-west-2.compute.internal:8042 0 ip-10-0-0-101.us-west-2.compute.internal:42205 LOST ip-10-0-0-101.us-west-2.compute.internal:8042 0 ip-10-0-0-103.us-west-2.compute.internal:60803 LOST ip-10-0-0-103.us-west-2.compute.internal:8042 0 ip-10-0-0-102.us-west-2.compute.internal:59201 LOST ip-10-0-0-102.us-west-2.compute.internal:8042 0 ===== (3) We rebooted NodeManager once again, and we can confirm information of NodeManager which was RUNNING in (2), but cannot confirm the information of NodeManager which was LOST in (2). [ec2-user@ip-10-0-0-4 ~]$ yarn node -list -all 13/10/24 18:50:04 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable 13/10/24 18:50:06 INFO client.RMProxy: Connecting to ResourceManager at ip-10-0-0-10.us-west-2.compute.internal/10.0.0.10:8032 Total Nodes:6 Node-Id Node-State Node-Http-Address Number-of-Running-Containers ip-10-0-0-103.us-west-2.compute.internal:37221 RUNNING ip-10-0-0-103.us-west-2.compute.internal:8042 0 ip-10-0-0-102.us-west-2.compute.internal:39940 RUNNING ip-10-0-0-102.us-west-2.compute.internal:8042 0 ip-10-0-0-101.us-west-2.compute.internal:36953 RUNNING ip-10-0-0-101.us-west-2.compute.internal:8042 0 ip-10-0-0-101.us-west-2.compute.internal:48281 LOST ip-10-0-0-101.us-west-2.compute.internal:8042 0 ip-10-0-0-103.us-west-2.compute.internal:51291 LOST ip-10-0-0-103.us-west-2.compute.internal:8042 0 ip-10-0-0-102.us-west-2.compute.internal:52128 LOST ip-10-0-0-102.us-west-2.compute.internal:8042 0 ===== We should make modifications in either following, a. we don't set LOST, if it is the reboot of the same server. In other words, LOST is not displayed in (2). b. we output all LOST information in (3).