Test below enables profiler on the shard and inserts 20,000 docs. Starts logging the following message on every insert after ~20,000 docs:
m30999| 2014-03-31T06:55:55.948-0400 [conn2] ChunkManager: time to load chunks for test.foo: 0ms sequenceNumber: 4 version: 1|4||533949b659411fe095434ebe based on: 1|2||533949b659411fe095434ebe m30999| 2014-03-31T06:55:55.948-0400 [conn2] autosplitted test.foo shard: ns: test.foo, shard: shard0000:localhost:30000, lastmod: 1|2||000000000000000000000000, min: { x: 1.0 }, max: { x: MaxKey } on: { x: 11466.0 } (splitThreshold 471859) (migrate suggested) m30000| 2014-03-31T06:55:56.333-0400 [conn9] warning: Caught Assertion while trying to profile insert against test.foo: 13388 [test.foo] shard version not ok in Client::Context: this shard contains versioned chunks for test.foo, but no version set in request ( ns : test.foo, received : 0|0||000000000000000000000000, wanted : 1|4||533949b659411fe095434ebe, send )
var st = new ShardingTest({shards: 1, mongos: 1}); var db = st.getDB("test"); st.adminCommand({enablesharding: "test"}); st.adminCommand({shardcollection : "test.foo", key: {x: 1}}); st.d0.getDB('test').setProfilingLevel(2); // Start a parallel shell to insert new documents on the primary. inserter = startParallelShell('for( i = 1; i < 25000; ++i ) { db.foo.save( { x:i } );}'); inserter(); printjson(st.printShardingStatus()); printjson(st.s0.getDB('test').getCollection('foo').stats());
- is duplicated by
-
SERVER-13414 profiling can throw stale config after write
- Closed