The following explains the procedure for handling (specifying) host names when IP addresses are inherited. The system configuration in this case is a 1-to-1 switchover configuration.
Item | Host name to be specified | |
---|---|---|
Host name specified in system common definition | pdunit -x hostA pdunit -x hostC pdstart -x hostA | |
Host name specified in unit control information definition | pd_hostname=host1 | |
Host name specified during rlogin | hostA | |
Host name specified in operation commands* | For single server | hostA (host name should be omitted) |
For utility special unit | hostC | |
Host name specified in utilities | For single server | hostA (host name should be omitted) |
For utility special unit | hostC | |
Host name displayed in messages | hostA | |
Host name displayed in statistical information | hostA | |
Host name specified in PDHOST operand of client environment definition | hostA |
* The unit identifier may be specified instead of the host name.
The following explains the procedure for handling (specifying) host names when IP addresses are not inherited. The system configuration in this case is a 1-to-1 switchover configuration.
Item | Host name to be specified | |
---|---|---|
Host name specified in system common definition | pdunit -x hostA -c hostB pdunit -x hostC pdstart -x hostA | |
Host name specified in unit control information definition | pd_hostname=host1 | |
Host name specified during rlogin | hostA or hostB (specify the host name of the running system) | |
Host name specified in operation commands1 | For single server | hostA (host name should be omitted) |
For utility special unit | hostC | |
Host name specified in utilities | For single server | hostA (host name should be omitted) |
For utility special unit | hostC | |
Host name displayed in messages | hostA | |
Host name displayed in statistical information | hostA | |
Host name specified in PDHOST operand of client environment definition | hostA and hostB2 |
1 The unit identifier may be specified instead of the host name.
2 Specify the host names of the primary system and secondary system in the PDHOST operand of the client environment definition. If you specify the host names in this manner, changing the value specified in the PDHOST operand will not be necessary, even after a system switchover. However, when the running system switches from the primary system to the secondary system, a UAP attempts to connect to the primary system (standby system). Because the UAP connects to the secondary system (running system) after the connection could not be established, the UAP processing time increases by the amount of time it takes to establish this new connection. To resolve this problem, set the host name for the client connection to an alias IP address and make sure that IP addresses are inherited. For details about specifying host names in such cases, see (3) IP addresses not inherited (only IP addresses for client connection inherited).
The following explains the procedure for handling (specifying) host names when the IP addresses HiRDB/Single Servers use are not inherited, but IP addresses for client connections are inherited. The system configuration in this case is a 1-to-1 switchover configuration.
Item | Host name to be specified |
---|---|
Host name specified in system common definition | pdunit -x hostA -c hostB pdstart -x hostA |
Host name specified in unit control information definition | pd_hostname=host1 |
Host name specified during rlogin | hostA or hostB (specify the host name of the running system) |
Host name specified in operation commands* | hostA (host name should be omitted) |
Host name specified in utilities | hostA (host name should be omitted) |
Host name displayed in messages | hostA |
Host name displayed in statistical information | hostA |
Host name specified in PDHOST operand of client environment definition | hostC |
* The unit identifier may be specified instead of the host name.