Troubleshooting

1. To troubleshoot problems with presto-admin or Presto, you can use the incident report gathering commands from presto-admin to gather logs and other system information from your cluster. Relevant commands:

   • collect logs
   • collect query_info
   • collect system_info

2. You can find the presto-admin logs in the /var/log/prestoadmin directory.

3. You can check the status of Presto on your cluster by using server status.

4. If Presto is not running and you try to execute any command from the Presto CLI you might get:

   $ Error running command: Server refused connection: http://localhost:8080/v1/statement

   To fix this, start Presto with:

   $ sudo ./presto-admin server start

5. If the Presto servers fail to start or crash soon after starting, look at the presto server logs on the Presto cluster /var/log/presto for an error message. You can collect the logs locally using collect logs. The relevant error messages should be at the end of the log with the most recent timestamp. Below are tips for some common errors:

   • Specifying a port that is already in use: Look at Presto Port Configuration to learn how to change the port configuration.
   • An error in a connector configuration file, such as a syntax error or a missing connector.name property: correct the file and deploy it to the cluster again using connector add

6. The following error can occur if you do not have passwordless ssh enabled and have not provided a password or if the user requires a sudo password:

   Fatal error: Needed to prompt for a connection or sudo password (host: master), but input would be ambiguous in parallel mode
   

   See SSH Configuration for information on setting up passwordless ssh and on providing a password, and Sudo Password Specification for information on providing a sudo password.