-
Type:
Improvement
-
Resolution: Unresolved
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: Shell
-
Labels:None
-
Query Optimization
-
(copied to CRM)
Say we have two documents with two fields in a collection (testing), with a compound unique index:
use test
db.testing.insert({ a: 1, b: 1 })
db.testing.insert({ a: 1, b: 2 })
db.testing.createIndex({ a: 1, b: 1 }, { unique: true })
The error message generated by a call to findAndModify that violates the unique index is as follows:
test> db.testing.findAndModify({ query: { a: 1,b: 1 }, update: { a: 1,b: 2 } })
2018-03-12T21:46:46.427-0400 E QUERY [thread1] Error: findAndModifyFailed failed: {
"operationTime": Timestamp(1520905600, 1),
"ok": 0,
"errmsg": "E11000 duplicate key error collection: test.testing index: a_1_b_1 dup key: { : 1.0, : 2.0 }",
"code": 11000,
"codeName": "DuplicateKey",
"$clusterTime": {
"clusterTime": Timestamp(1520905600, 1),
"signature": {
"hash": BinData(0, "eiANq76RZ704EWnDJSY3orlIWxM="),
"keyId": NumberLong("6532150305184808961")
}
}
} :
_getErrorWithCode@src/mongo/shell/utils.js:25:13
DBCollection.prototype.findAndModify@src/mongo/shell/collection.js:724:1
@(shell):1:1
In the above error message, codeName is specified. However in an insert or update operation, the codeName field is not specified:
test> db.testing.insert({a:1,b:1})
E11000 duplicate key error collection: test.testing index: a_1_b_1 dup key: { : 1.0, : 1.0 }
WriteResult({
"nInserted": 0,
"writeError": {
"code": 11000,
"errmsg": "E11000 duplicate key error collection: test.testing index: a_1_b_1 dup key: { : 1.0, : 1.0 }"
}
})
These error messages should be made consistent across all methods in the shell.
- is related to
-
SERVER-36755 writeErrors codeName is set in sharded cluster topology only
-
- Backlog
-