![[Contents]](FIGURE/CONTENT.GIF)
Nonstop Database, HiRDB Version 9 System Operation Guide
![[Index]](FIGURE/INDEX.GIF)
![[Back]](FIGURE/FRONT.GIF)
![[Next]](FIGURE/AFTER.GIF)
If an error occurs in a HiRDB server process, HiRDB terminates that server process abnormally. In this case, HiRDB starts a transaction recovery process and rolls back each server. Since the applicable server process is automatically restarted afterwards, there is no action that the HiRDB administrator needs to take.
Depending on the extent of the error's impact, HiRDB might terminate abnormally the unit on which the server process was being executed. In such a case, the HiRDB administrator must restart the abnormally terminated unit (although the unit might be restarted automatically, depending on the specification of the pd_mode_conf operand).
The following explains the actions that the HiRDB administrator must take when abnormal termination of a server process causes abnormal termination of a unit.
The following table shows the action to be taken by the HiRDB administrator in the event that a server in a HiRDB single server configuration terminates abnormally.
Table 20-2 HiRDB administrator's action in the event of an abnormal termination of a server in a HiRDB single server configuration
| HiRDB processing | HiRDB administrator's action |
|---|---|
| Terminates the server in a HiRDB single server configuration abnormally | Eliminate the cause of the error and restart the server in a HiRDB single server configuration (the server in a HiRDB single server configuration might be restarted automatically, depending on the pd_mode_conf operand specification), then re-execute the previous processing. |
The following table shows the actions to be taken by the HiRDB administrator in the event that a server in a HiRDB parallel server configuration terminates abnormally.
Table 20-3 HiRDB administrator's actions in the event of an abnormal termination of a server in a HiRDB parallel server configuration
| Error status | HiRDB processing | HiRDB administrator's action |
|---|---|---|
| Process error at unit controller | Terminates abnormally the unit where the corresponding process is located. | Restart the corresponding unit (the unit might be restarted automatically, depending on the pd_mode_conf operand specification). |
| Error in system manager process | Terminates abnormally the unit where the system manager is located. HiRDB does not accept any operation command or UAP execution request while the system manager is shut down. No message log can be collected during that time. | Eliminate the cause of the error and restart the corresponding unit (the unit might be restarted automatically, depending on the pd_mode_conf operand specification), then re-execute the previous processing. |
| Error in front-end server process | Terminates abnormally the unit where the front-end server is located. HiRDB does not accept any processing request while the front-end server is shut down. It treats a processing request as a timeout-level communication error. If a front-end server process terminates abnormally, HiRDB recovers the process. If it is necessary to shut down the unit in such a case, HiRDB terminates abnormally the unit where the corresponding process is located. |
Eliminate the cause of the error and restart the corresponding unit (the unit might be restarted automatically, depending on the pd_mode_conf operand specification), then re-execute the previous processing. |
| Error in dictionary server process | Terminates abnormally the unit where the dictionary server is located. | |
| Error in back-end server process | Terminates abnormally the unit where the back-end server is located. |
All Rights Reserved. Copyright (C) 2011, 2015, Hitachi, Ltd.