mongos> sh.shardCollection("mf.mb_device_login_log",{"_plat" : 1,"_uid" : 1,"_tm" : 1})
{ "code" : 20, "ok" : 0, "errmsg" : "sharding not enabled for db mf" }

檢視索引,索引存在

mongos> db.mb_device_login_log.getIndexes()
[
    {
        "v" : 2,
        "key" : {
            "_plat" : 1,
            "_uid"
 
 : 1,
            "_tm" : 1
        },
        "name" : "_plat_1__uid_1__tm_1",
        "ns" : "mf.mb_device_login_log"
    }
]

對db啟用分片，分片已經存在

mongos> sh.enableSharding("mf")
{
"code" : 23,
"ok" : 0,
"errmsg" : "sharding already enabled for database mf"
}

問題分析，懷疑是上次分片後，配置資料沒有及時重新整理
重新整理配置資料

mongos> db.adminCommand("flushRouterConfig")
{ "flushed" : true, "ok" : 1 }

這次執行分片成功

mongos> sh.shardCollection("mf.mb_device_login_log",{"_plat" : 1,"_uid" : 1,"_tm" : 1})
{ "collectionsharded" : "mf.mb_device_login_log", "ok" : 1 }
mongos>