CODEWITHSUNDEEP
mongodb
chaliasos
chaliasos

Reputation: 9783

Cannot query a shard collection

I created a shard with two replica sets. Each replica set have 3 mongod server and they work fine. Also, the mongos server seems to work. Here is it's status:

mongos> sh.status()
--- Sharding Status ---
  sharding version: {
        "_id" : 1,
        "version" : 3,
        "minCompatibleVersion" : 3,
        "currentVersion" : 4,
        "clusterId" : ObjectId("51c5be46dda9b0a5cd83ad64")
}
  shards:
        {  "_id" : "s1",  "host" : "s1/RemotePC_1:27017,RemotePC_1:27018,RemotePC_1:27019" }
        {  "_id" : "s2",  "host" : "s2/RemotePC_2:27017,RemotePC_2:27018,RemotePC_2:27019" }
  databases:
        {  "_id" : "admin",  "partitioned" : false,  "primary" : "config" }
        {  "_id" : "student",  "partitioned" : true,  "primary" : "s1" }
                student.grades
                        shard key: { "_id" : 1 }
                        chunks:
                                s1      1
                        { "_id" : { "$minKey" : 1 } } -->> { "_id" : { "$maxKey" : 1 } } on : s1 { "t" : 1, "i" : 0 }

But, when I run any command (insert, find, count, mongoimport) on the student.grades collection I get back this error:

mongos> db.grades.count()
Sat Jun 22 20:07:00.968 JavaScript execution failed: count failed: {
        "code" : 10429,
        "ok" : 0,
        "errmsg" : "exception: setShardVersion failed host: RemotePC_2:27017 { oldVersion: Timestamp 0|0, oldVersionEpoch: ObjectId
        ('000000000000000000000000'), errmsg: \"exception: all servers down/unreachable when querying: My-PC:57017,My-PC:57018,My-PC:57019\", code: 8002, ok: 0.0 }"
} at src/mongo/shell/query.js:L180

Why? Am I missing something?

Upvotes: 1

Views: 3297

Answers (1)

Asya Kamsky
Asya Kamsky

Reputation: 42352

All the members of the sharded cluster must be able to reach all the other members.

You are running mongos on your local machine, it can reach config servers which also run on your local machine, and it can reach the shards which are running on remote machines.

However, if the shards on remote machines cannot reach the config servers on your local machine they will not be able to operate properly within the cluster (and that is the error message that you are seeing).

The solution to this would be to restart mongos with the config server argument (--configdb) specified as hostnames which are routable (aka findable) from "remote-PC". If you are running multiple mongos processes you will need to do that for all of them.

Upvotes: 2

Related Questions