HVR Installation Location

The HVR software may be installed as an HVR hub and as a remote agent.

For using HVR as HVR hub, the HVR software must be installed on the hub machine. If a database involved in replication is not located on the hub machine, HVR can connect to it in two ways: either using the network database's own protocol (e.g. SQL*Net or Ingres/Net) or using an HVR remote connection (recommended).

HVR remote agent is an additional installation of HVR (within the HVR distributed architecture) that does not play the role of the HVR hub and does not run the Scheduler service. The HVR remote agent is installed on a remote machine of the source or target location for optimized communication with the HVR hub. For the HVR remote agent location, the HVR software can also be installed on the remote machine to connect to the hub machine. If the DBMS's own protocol is used for remote connection, then no HVR software needs be installed on the remote agent machine.

If the HVR replication configuration is altered, all the reconfiguration steps are performed on the hub machine; installation files on the remote machine are unchanged.

It is recommended that HVR is installed under its operating system account (e.g. hvr) on the hub machine and each remote location containing the replicated database or directory. However, HVR can also use the account that owns the replicated tables (e.g. the DBA's account) or it can use the DBMS owner account (e.g. Oracle or Ingres). Any login shell is sufficient.

HVR makes no assumptions about there being only one installation per machine, and locations sharing a single machine need only have HVR installed once and HVR can easily replicate data between these locations.