DWS failover

Certain DWS clients have implemented support for DWS failover and DCE failover, to reduce device downtime. An administrator can configure up to four DWS nodes and up to four DCE nodes. DWS and DCE do not need to be on the same machine. Currently DWS failover is only available for the Unified Client for Brother and the Unified Client for HP.

DWS failover has a single prime DWS and up to three backup nodes. If DWS 1 fails, then devices will fail down the chain from DWS 1 to DWS 2 to DWS 3 to DWS 4, until the device is offline. The device can only revert back to DWS 1, even if DWS 2 or DWS 3 is online before DWS 1. This behavior reduces the amount of device reprogramming. Each time a device changes DWS, it is reprogrammed, which might result in some devices rebooting. However, a device cannot revert back to DWS 1 if a user is actively using that device, to ensure that any current operation has a chance to complete. For setup examples, see DWS failover scenarios.

Actions at the device during a failover scenario might be lost if the DWS goes offline and cannot respond to scan uploads or Web page requests. Users must log out of a device during failover as the backup DWS might not be able to complete reconfiguration.

DCE failover

DCE failover for the Unified Client for Brother and the Unified Client for HP is similar to DCE failover on other embedded Unified Clients. During failover from DCE 1 to a backup DCE, the current device user is logged out, and must then log in again to continue using the system. Any work in progress might be lost during the transition, however a best effort is made to allow scans or other actions to complete.

Unlike DWS failover, DCE always reverts to the highest DCE that is currently online. If a system points to DWS 1, which in turn points to DCE 4, and DCE 3 comes back online, the device continues to point to DWS 1 but changes over to DCE 3 once the current session is complete. This revert action is only done when a user is logged out to ensure that current actions are completed.

  • If your environment uses DWS failover and DCE failover, you do not need to keep both nodes in sync. For example, DWS 4 can point to DCE 1 or DWS 1 can point to DCE 4.
  • If all components are not online for your failover setup, you might see warning messages even though everything appears to be working. It is recommended that your full system is online during DRS configuration changes or when adding or removing devices.
  • You can use the Update Configuration action to set up a previously offline DWS node with the failover configuration. You cannot change failover configuration this way, only push the existing failover configuration to a node that was offline during the initial device installation.
  • DWS failover is not supported behind a network load balancer. However, Equitrac and AutoStore servers can be behind a network load balancer with DWS nodes set up for failover.

Failover is not instant. You might experience a login failure while a DWS or DCE failover occurs. As there is no shared data, transactions might be lost, or reported against an unidentified user instead of the user that triggered the workflow.