Seems like the same issue that was present in jstests/sharding/shard5.js - the second mongos probably doesn't see the movePrimary:
[js_test:shard3] 2015-10-07T06:29:44.690+0000 s20015| 2015-10-07T06:29:44.689+0000 I COMMAND [conn1] movePrimary dropping database on ip-10-144-96-213:20010, no sharded collections in test [js_test:shard3] 2015-10-07T06:29:44.691+0000 s20015| 2015-10-07T06:29:44.690+0000 I SHARDING [conn1] about to log metadata event: { _id: "ip-10-144-96-213-2015-10-07T06:29:44.690+0000-5614bbd8fd609a730739abae", server: "ip-10-144-96-213", clientAddr: "127.0.0.1:41563", time: new Date(1444199384690), what: "movePrimary", ns: "test", details: { database: "test", from: "ip-10-144-96-213:20010", to: "shard0001:ip-10-144-96-213:20011", shardedCollections: [] } } ... [js_test:shard3] 2015-10-07T06:29:45.356+0000 s20016| 2015-10-07T06:29:45.354+0000 D SHARDING [Balancer] found 1 shards listed on config server(s) [js_test:shard3] 2015-10-07T06:29:45.357+0000 2015-10-07T06:29:45.354+0000 E QUERY [thread1] Error: [3] != [0] are not equal : other A : [js_test:shard3] 2015-10-07T06:29:45.357+0000 doassert@src/mongo/shell/assert.js:15:14 [js_test:shard3] 2015-10-07T06:29:45.357+0000 assert.eq@src/mongo/shell/assert.js:43:5 [js_test:shard3] 2015-10-07T06:29:45.357+0000 @jstests/sharding/shard3.js:36:1 [js_test:shard3] 2015-10-07T06:29:45.357+0000 [js_test:shard3] 2015-10-07T06:29:45.357+0000 failed to load: jstests/sharding/shard3.js
- related to
-
SERVER-20690 Test with multiple mongos can fail because one of them talked to a stale config server secondary
- Closed