Hitachi

Hitachi Advanced Database Command Reference


39.6 Notes

  1. If the previous termination mode was abnormal termination, the command displays the KFAA91104-Q message when the HADB server is started. This message asks if the existing troubleshooting information can be deleted. If you have already used the adbinfoget command to collect troubleshooting information, or if there is no need to collect troubleshooting information, enter Y as the response to the KFAA91104-Q message. When Y is entered, the command discards the troubleshooting information and starts the HADB server. If you have not collected troubleshooting information with the adbinfoget command, enter N as the response to the KFAA91104-Q message. When N (or anything other than Y) is entered, the adbstart command's processing is canceled.

  2. If you execute the adbstart command after the HADB server has been upgraded, the KFAA91107-Q message asking whether the database is to be upgraded will be output. If it is safe to upgrade the database, enter Y. If the adbstart command terminates normally, the HADB server has been upgraded. If upgrading might cause a problem, enter N (or any character other than Y), and the HADB server upgrade processing will be canceled and the adbstart command will terminate.

    For details about how to upgrade HADB servers, see Upgrading the HADB server version in Building a System in the HADB Setup and Operation Guide.

  3. If the audit trail facility was enabled the last time the HADB server terminated, the audit trail facility remains in effect the next time the HADB server is started. The specification of the --write-error option of the adbaudittrail command is also inherited. For details about the adbaudittrail command, see 3. adbaudittrail (Manage the Audit Trail Facility).

  4. If the audit trail facility is enabled, a new current audit trail file is created when the HADB server is started. Note that when the HADB server is restarted, the current audit trail file used the last time the HADB server terminated is renamed. Then, a new current audit trail file is created.

  5. If the audit trail facility is enabled, HADB server cannot be started unless all the following conditions are satisfied. If the conditions are not satisfied, the adbstart command results in an error.

    • An existing directory is specified for the adb_audit_log_path operand in the server definition.

    • Read, write, and execution privileges are set so that the HADB administrator can access the directory specified for the adb_audit_log_path operand in the server definition.

    • The number of audit trail files stored under the directory specified for the adb_audit_log_path operand in the server definition is equal to or smaller than the value specified for the adb_audit_log_max_num operand in the server definition.

  6. To return a node that has been disconnected from the multi-node configuration to the running multi-node configuration while the multi-node function is being used, perform the following:

    • Make sure there are no ongoing transactions. For example, use the adbchgsrvmode command to change the operation mode of the HADB server on the master node to maintenance mode. For details about the adbchgsrvmode command, see 9. adbchgsrvmode (Change the HADB Server Operation Mode).

    • Check the status of HADB servers on other nodes in advance. If an HADB server on any other node is engaged in termination processing, the disconnected node cannot be returned to the multi-node configuration.

      To check the status of the HADB servers, execute the adbls -d srv command on each node. On the STATUS column in the output results, you can check the status of the HADB server on the node on which the command was executed. For details about the adbls -d srv command, see 30. adbls -d srv (Display the HADB Server Status).

    • If you want to return multiple nodes to the multi-node configuration, return them one by one. You cannot return multiple nodes at the same time.

    For details about returning a node to the multi-node configuration, see Returning a node to the multi-node configuration in Operations when a node failure occurs in Operations When Using the Multi-Node Function in the HADB Setup and Operation Guide.