We are using ES 1.7.1 on one server, on another server, we have version 2.3.3 installed (latest up to date).
JVM version:: 1.8.0_66 on Ubuntu 14.04.3 LTS
For both servers, we have index.number_of_shards set to 1 in /etc/elasticsearch/elasticsearch.yml
On ES 1.7.1, GET _/cat/allocation?v yields
shards disk.used disk.avail disk.total disk.percent host ip node
38 24.4gb 91.3gb 115.7gb 21 elastic171 192.168.2.171 ns171
38 UNASSIGNED
As for ES 2.3.3, the same command yields
shards disk.indices disk.used disk.avail disk.total disk.percent host ip node
41 13.9gb 21.9gb 93.8gb 115.7gb 18 192.168.2.104 192.168.2.104 ns0
41 UNASSIGNED
Why is it showing too many shards (38 and 41 above) ?