Replace a failed YB-TServer

If you have a failed YB-TServer in a YugabyteDB cluster that needs to be replaced, follow these steps:

Start the new YB-TServer

Install and then start a new YB-TServer, making sure it is in the same placement group as the one you are replacing.

For details on starting YB-TServers and more options, see Start YB-TServers.

Blacklist the failed YB-TServer

To blacklist the failed YB-TServer, run the following command:

For details, see change_blacklist command in the yb-admin reference page.

Wait for the rebalance to complete

Wait for the data to drain from the failed YB-TServer and for the data to be loaded into the new one. You can check for the completion of rebalancing by running the following command:

For details on using this command, see get_load_move_completion.

Kill the failed YB-TServer

When the data move is complete (100%), kill the failed YB-TServer by stopping the yb-tserver process, or terminating the VM.

Now, wait for the YB-TServer to be marked as DEAD by the YB-Master leader. The YB-Master leader will mark the server as DEAD after not responding for one minute (based on the tserver_unresponsive_timeout_ms (default is 60000).

To verify that the failed YB-TServer is dead, open your web browser to $MASTER_LEADER_IP:7000/tablet-servers and check the output.

Remove the failed YB-TServer from the blacklist

Now that the replacement YB-TServer is up and running, and loaded with data, remove the address for the failed YB-TServer from the blacklist.