elasticsearch: All primary shards become inactive when restart a single node cluster -

i started new elasticsearch cluster , added new index, shut down. when try start again there no active shards more. shards became inactive , unassigned. use index settings below:

"settings": {             "number_of_shards": 32,             "number_of_replicas": 3         } 

here health output:

{   "cluster_name" : "sailcraft",   "status" : "red",   "timed_out" : false,   "number_of_nodes" : 1,   "number_of_data_nodes" : 1,   "active_primary_shards" : 0,   "active_shards" : 0,   "relocating_shards" : 0,   "initializing_shards" : 0,   "unassigned_shards" : 128,   "number_of_pending_tasks" : 0 } 

here same problem has been marked solved. don't think that's solution.

edit:

it's same question think post not right answer. said need 2 more nodes if have 2 more replica shards. read related docs , there not such restrictions.

edit2: solution: setting index.recovery.initial_shards provides following: basically, when using local gateway, "shard" recovered once quorum of copies found among nodes in cluster. lets take n shard, 1 replica (2 copies each shard). default (quorum) means once singe copy of shard found, recovered. if have 2 replicas (3 copies), recover shard once 2 copies found. can set setting, in case lost large number of nodes, , quorum strict setting. can set on "live" index (which in "red" state, obviously, because not shards recovered).

here's the documentation mentioning this. indeed, it's not evident, it's there:

index.recovery.initial_shards

when using local gateway particular shard recovered if there can allocated quorum shards in cluster. can set to:

quorum (default)

quorum-1 (or half)

full

full-1.

number values supported, e.g. 1.

if want have node running, set index.recovery.initial_shards: 1 in elasticsearch.yml file.