Nonstop Database, HiRDB Version 9 System Operation Guide
When the automatic log unloading facility is used, HiRDB unloads automatically the system log files that are waiting to be unloaded. Unload log files are created under a directory created in advance by the HiRDB administrator (this directory is called the directory for unload log files). The following figure shows the operation of the automatic log unloading facility.
Figure 3-6 Automatic log unloading facility
The automatic log unloading facility stops operating when unload log files cannot be created in the directory for unload log files because of a full disk, disk error, or similar condition. To prevent this from happening, the HiRDB administrator must transfer the unload log files in the directory for unload log files to a separate medium on a regular basis. At the same time, the HiRDB administrator deletes unneeded unload log files.
When multiple directories for unload log files are created, the automatic log unloading facility stops only when unload log files cannot be created in any of the directories.
The automatic log unloading facility cannot be used when HiRDB is not set to check the unload status of system log files (when pd_log_unload_check = N is specified).
The format of the names of the unload log files that are created by the automatic log unloading facility is shown below (the names of created unload log files are displayed in the KFPS01212-I message):
server-name_server-run-ID-and-generation-number_file-group-name
The names of unload log files created by the automatic log unloading facility can be checked by entering the pdlogatul command.
When there are multiple unload log file directories, unload log files that cannot be created in one directory because of a full disk, disk error, or similar condition, are created in another directory. HiRDB selects the directories to be used in the order of their specification in the pd_log_auto_unload_path operand.
For example, if pd_log_auto_unload_path = "Directory 1","Directory 2","Directory 3",... is specified, HiRDB uses Directory 1 first, then Directory 2, then Directory 3, and so on. However, if Directory 2 is not available because of a disk error, or similar condition, Directory 3 is used after Directory 1.
When HiRDB is started normally, Directory 1 is used first, assuming that there are no unload log files in Directory 1. If Directory 1 contains unload log files, Directory 2 is used.
When HiRDB is restarted (including during system log file swapping), the directory used is the one that was being used when HiRDB was terminated.
All Rights Reserved. Copyright (C) 2011, 2015, Hitachi, Ltd.