Crash due to race between DCP Consumer message processing and stream shutdown

5.5.4-4334 - centos longevity - following memcached exit observed:

Service 'memcached' exited with status 139. Restarting. Messages:
2019-04-11T11:19:07.686740-07:00 WARNING (DISTRICT) Slow scheduling for NON_IO task 'Adjusting hash table sizes.' on thread nonIO_worker_0. Schedule overhead: 11 s
2019-04-11T11:19:07.686887-07:00 WARNING (HISTORY) Slow scheduling for NON_IO task 'Adjusting hash table sizes.' on thread nonIO_worker_0. Schedule overhead: 11 s
2019-04-11T11:19:07.687010-07:00 WARNING (ITEM) Slow scheduling for NON_IO task 'Adjusting hash table sizes.' on thread nonIO_worker_0. Schedule overhead: 11 s
2019-04-11T11:19:07.711580-07:00 WARNING (HISTORY) Slow runtime for 'Checkpoint Remover on vb 209' on thread nonIO_worker_2: 16 ms
2019-04-11T11:19:07.722746-07:00 WARNING 136: (HISTORY) DCP (Consumer) eq_dcpq:replication:ns_1@172.23.96.148->ns_1@172.23.96.56:HISTORY - vb:267 Got error 'no memory' while trying to process mutation with seqno:3116
2019-04-11T11:19:07.722998-07:00 WARNING (HISTORY) DCP (Consumer) eq_dcpq:replication:ns_1@172.23.97.239->ns_1@172.23.96.56:HISTORY - (vb 623) End stream received but no such stream for this vBucket

Logs: http://supportal.couchbase.com/snapshot/8ff7d91dd5a771c1904307a1f2f35743::0
Cluster: 172.23.108.103
console: http://172.23.109.231/job/centos-systest-launcher/1767/console