Fail over to a second PACS

Goal: when the primary PACS (or its storage node) stops answering, studies go to a second one, and return to the primary once it is back.

Steps

1. Destinations: PACS A (primary) and PACS B (standby), each with its own address and AE title.
2. Group: Configuration › Destinations › Add destination, type Group, name PACS, Mode Failover, Members PACS A then PACS B. Failover after (s): how long the primary may be unreachable before studies move on (default 60).
3. Routes: send to the group PACS instead of PACS A.

What happens

• While PACS A answers, everything goes to it.
• When it has been unreachable for the failover delay, new studies go to PACS B, and instances already queued for PACS A through the group move to PACS B too. The history shows them as Rerouted.
• The router checks PACS A with C-ECHO every 30 s; once it answers, traffic returns to it.

Balancing instead

To spread studies over several servers (AI servers, archive nodes), choose Mode Balance: each study goes to one member, whole.

See Destination groups.