There is a possibility to start cluster in a "local" mode, serving data from one working node exclusively.
Once your DSS node-b [or node-a, for that matter] goes down, working node will automatically switch itself into "source" mode, changing its
volumes state into "source" as well. Also. cluster status will become "degraded", changing into "local" mode and serving files
If - for some reason - cluster is stopped, there is still possibility to run it in "local" mode.
Just click on a "Start" button on a "Failover" section and confirm that you would like to run cluster in "local" mode.
All connections should then be redirected to the working node you are operating on, so data will become accessible right away.
Once the issues with other node of the cluster is solved, cluster will automatically return to auto failover, "Running - OK" mode.
NOTE: please be aware that modifying other settings of the cluster might prohibit you from running cluster in "local" mode
(for example: removing any of the replication tasks - but NOT stopping them)