Xdcr - Replication now* halts on *rebooted node* only, on destination cluster.

Setup
-Create unidirectional replication between 2 clusters[1:5 nodes] on a saslbucket
-Load 2M items on source

• Reboot one node on destination cluster.

-Error
Replication from source cluster halts.

Seeing these crash reports on the source logs [This is likely from an older crash ... But I dont see any new crash reports on the source logs apart from these.]
"could not open http://Administrator:*****@10.3.121.37:8092/saslbucket%2f64%3b7f5644eb7aaa4a015c601ff7d0659d72/">>}
=========================CRASH REPORT=========================

• crasher:
• initial call: xdc_vbucket_rep:init/1
• pid: <0.10569.1>
• registered_name: []
• exception exit: {db_not_found,<<"could not open http://Administrator:*****@10.3.121.37:8092/saslbucket%2f84%3b7f5644eb7aaa4a015c601ff7d0659d72/">>}
• in function gen_server:init_it/6
• ancestors: [<0.8393.0>,<0.8389.0>,xdc_replication_sup,ns_server_sup,
• ns_server_cluster_sup,<0.60.0>]
• messages: [src_db_updated]
• links: [<0.10570.1>,<0.8393.0>]
• dictionary: []
• trap_exit: false
• status: running
• heap_size: 46368
• stack_size: 24
• reductions: 5873
• neighbours:
• neighbour: [ {pid,<0.10570.1>}

  ,

• {registered_name,[]}

  ,

• {initial_call,{lhttpc_manager,init,['Argument__1']}},
• {current_function,{gen_server,loop,6}},
• {ancestors,[<0.10569.1>,<0.8393.0>,<0.8389.0>, - xdc_replication_sup,ns_server_sup, - ns_server_cluster_sup,<0.60.0>]}

  ,

• {messages,[]}

  ,

• {links,[<0.10569.1>,#Port<0.66678>]}

  ,

• {dictionary,[]}

  ,

• {trap_exit,false}

  ,

• {status,waiting}

  ,

• {heap_size,377}

  ,

*Build 1611, has the fix from http://www.couchbase.com/issues/browse/MB-6324

Adding logs from the cluster.

If this is something we dont support for XDCR, then we should document this.
We see this w/ any topology change on destination cluster, replication halts.