Multiple StorMan servers

You can install StorMan on multiple Management Units (MUs), define these in the SHC-OSD parameter file, and use them dynamically in rotation.
This is particularly important when stringent requirements are made for availability (High Availability, HA) or for configurations for disaster recovery (DR).

Figure 6: Two StorMan servers in an HA configuration

However, only one of the defined StorMan servers is ever active. When it starts, SHC-OSD activates the first StorMan server defined in the SHC-OSD parameter file which is in the OK status (StorMan server is running). Information on the active StorMan servers can be obtained with the /SHOW-SHC-ENVIRONMENT INFORMATION=*STD command.

Information on all StorMan servers defined in the SHC-OSD parameter file can be obtained with the /SHOW-SHC-PROCESSING and /SHOW-SHC-ENVIRONMENT INFORMATION=*ALL commands.

The active StorMan server can be switched over automatically or manually during ongoing SHC-OSD operation:

• Automatic switchover

  SHC-OSD detects that the active StorMan server has crashed when the active StorMan server fails to respond to either the regular polling (default interval: 10 min.) or to an SHC-OSD command.

  In this case SHC-OSD switches to the first available StorMan server from the list of defined StorMan servers in the SHC-OSD parameter file.

  The switchover of the StorMan server is logged on the console with the messages NDE1836 (StorMan server no longer connected) and NDE1835 (Storman server connected). Message NDE1836 must be confirmed by the operator.
  

• Manual switchover by command

  The /MODIFY-SHC-ENVIRONMENT command enables you to switch over to any active StorMan server contained in the list of defined StorMan servers in the SHC-OSD parameter file during ongoing SHC-OSD operation.

  The switchover can also take place from a procedure.