FAQ on deployment

This topic describes frequently asked questions about OCP deployment.

Deploy OCP by using a CLI tool

Q1: The install temp ocp step timed out when I deploy OCP by using a CLI tool. What can I do?

A: Contact OceanBase technical support to obtain the latest OAT version.

Q2: An error occurred when I was deploying OCP. Error message: error tmp_ocp docker on xx.xx.xx.xx is NOT started . How do I solve the problem?

A: This problem may be caused by an outdated OAT version. We recommend that you contact OceanBase technical support to obtain the latest OAT version.

Q3: An RPC error occurred when I was adding a MetaDB host during the OCP deployment by using a CLI tool. Error message: Invalid token . What can I do?

A: Run the ps -ef | grep pos_proxy command to check for residual pos_proxy processes on the physical server and use the kill -9 PID command to clean them up if any.

Q4: By default, backup is disabled during OCP deployment. During a POC backup and recovery test, do I need to stop the running OCP before modifying the uster.conf file of the OBCL and executing install.sh -c 7 -i 7 ?

A: Yes. Perform the following steps:

1. Suspend a running OCP.

2. In the backup directory, modify the value of BACKUP_ENABLE to TRUE.

3. Run install.sh -c 7 -i 7.