There is no primary, how to make secondary as primary by force

[Avise Sudhakar Rao]

I needd your quick help

please respond

I have one serve in down

and another server in secondary state

but there is no primary

how can I make secondary as Primary forcely

?

please help

 

[Scott Hernandez]

You can do a reconfigure with a higher vote count so that you have a
majority of votes with the members which are still up. In general you
should never have an even number, so having 3 data members + 1 arbiter
is a bad idea.

http://docs.mongodb.org/manual/tutorial/reconfigure-replica-set-with-unavailable-members/
http://docs.mongodb.org/manual/reference/method/rs.reconfig/

> --
> --
> You received this message because you are subscribed to the Google
> Groups "mongodb-user" group.
> To post to this group, send email to mongod...@googlegroups.com
> To unsubscribe from this group, send email to
> mongodb-user...@googlegroups.com
> See also the IRC channel -- freenode.net#mongodb
>
> ---
> You received this message because you are subscribed to the Google Groups
> "mongodb-user" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to mongodb-user...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Avise Sudhakar Rao]

currently it is in monit:STARTUP2 mode

how can I set

[Avise Sudhakar Rao]

monit:STARTUP2> cfg = rs.conf()
{
        "_id" : "monit",
        "version" : 21,
        "members" : [
                {
                        "_id" : 0,
                        "host" : "172.25.122.181:27017"
                },
                {
                        "_id" : 2,
                        "host" : "172.25.122.20:27017",
                        "arbiterOnly" : true
                },
                {
                        "_id" : 3,
                        "host" : "172.25.122.182:27017"
                },
                {
                        "_id" : 4,
                        "host" : "172.25.122.186:27017"
                }
        ]
}
monit:STARTUP2> cfg.members[0].priority = 0.5
0.5
monit:STARTUP2> cfg.members[2].priority = 0.5
0.5
monit:STARTUP2> cfg.members[3].priority = 1
1
monit:STARTUP2> cfg.members[4].priority = 1
Fri Feb  8 20:52:57 TypeError: cfg.members[4] has no properties (shell):1
monit:STARTUP2> rs.reconfig(cfg,{force:true})
{
        "errmsg" : "exception: need most members up to reconfigure, not ok : 172.25.122.186:27017",
        "code" : 13144,
        "ok" : 0
}

I'm getting above error, please to make 172.25.122.182 as primary

[Asya Kamsky]

Dear Sudhakar:

In order to have a node act as Primary it has to get votes from majority of the cluster.

Because you have four members on the replica set, majority is 3.  If you reduce the replica set configuration to three then it will only take two votes to be a primary.

You say you want to make 172.25.122.182 as Primary - can you log into that server with "mongo --host 171.25.122.186" ?

Earlier you posted:

monit:RECOVERING> db.printSlaveReplicationInfo()
source:   172.25.122.181:27017
         syncedTo: Fri Feb 08 2013 12:14:55 GMT+0000 (UTC)
                 = 16998 secs ago (4.72hrs)
source:   172.25.122.20:27017
         no replication info, yet.  State: ARBITER
source:   172.25.122.186:27017
         syncedTo: Fri Feb 08 2013 13:49:48 GMT+0000 (UTC)
                 = 11305 secs ago (3.14hrs)
monit:RECOVERING> 

Since 172.25.122.182 is the machine you want to make primary, connect to it via mongo shell and do 'rs.reconfigure' command there.

If you remove the stalest replica from the configuration, the arbiter and 172.25.186.182 will be able to vote it back into primary status,

hopefully then your other secondary can catch up.  

Create a config which looks like this:

newconfig = {

        "_id" : "monit",
        "version" : 21,
        "members" : [
                {

                        "_id" : 2,
                        "host" : "172.25.122.20:27017",
                        "arbiterOnly" : true
                },
                {
                        "_id" : 3,
                        "host" : "172.25.122.182:27017"
                },
                {
                        "_id" : 4,
                        "host" : "172.25.122.186:27017"
                }
        ]
}

run rs.reconfig(newconfig, {force:true}) ON 172.25.122.186:27017

Once you've restored basic functionality, I would recommend investing time into setting up a more robust cluster, with larger oplogs

and always an odd number of members.

Asya

[Avise Sudhakar Rao]

Actually there is no data on 172.25.122.186

and if I start .186 then it will become primary and I will lost the data on .182 which is having production data

How I can run on .182 instead of .186?

Please help

[Asya Kamsky]

I'm sorry, I must have mis-pasted.

Change your config object to be the same but run the rs.reconfig(newconfig) command when connected to 172.25.122.182:27017

You can add higher priority to it in the config as well like this:

newconfig = {

        "_id" : "monit",
        "version" : 21,
        "members" : [
                {

                        "_id" : 2,
                        "host" : "172.25.122.20:27017",
                        "arbiterOnly" : true
                },
                {
                        "_id" : 3,
                        "host" : "172.25.122.182:27017",

                       "priority": 2

                },
                {
                        "_id" : 4,
                        "host" : "172.25.122.186:27017"
                }
        ]
}

On Friday, February 8, 2013 1:21:19 PM UTC-8, Sudhakar wrote:

Actually there is no data on 172.25.122.186

and if I start .186 then it will become primary and I will lost the data on .182 which is having production data

How I can run on .182 instead of .186?

Please help

On Sat, Feb 9, 2013 at 2:42 AM, Asya Kamsky <as...@10gen.com> wrote:

Dear Sudhakar:

In order to have a node act as Primary it has to get votes from majority of the cluster.

Because you have four members on the replica set, majority is 3.  If you reduce the replica set configuration to three then it will only take two votes to be a primary.

You say you want to make 172.25.122.182 as Primary - can you log into that server with "mongo --host 171.25.122.182" ?

run rs.reconfig(newconfig, {force:true}) ON 172.25.122.182:27017

[Avise Sudhakar Rao]

monit:SECONDARY> rs.status()
{
        "set" : "monit",
        "date" : ISODate("2013-02-08T21:39:41Z"),
        "myState" : 2,
        "members" : [
                {
                        "_id" : 2,
                        "name" : "172.25.122.20:27017",
                        "health" : 1,
                        "state" : 7,
                        "stateStr" : "ARBITER",
                        "uptime" : 72,
                        "lastHeartbeat" : ISODate("2013-02-08T21:39:41Z"),
                        "pingMs" : 1
                },
                {
                        "_id" : 3,
                        "name" : "172.25.122.182:27017",
                        "health" : 1,
                        "state" : 2,
                        "stateStr" : "SECONDARY",
                        "uptime" : 72,
                        "optime" : Timestamp(1360354594000, 148),
                        "optimeDate" : ISODate("2013-02-08T20:16:34Z"),
                        "self" : true
                },
                {
                        "_id" : 4,
                        "name" : "172.25.122.186:27017",
                        "health" : 0,
                        "state" : 8,
                        "stateStr" : "(not reachable/healthy)",
                        "uptime" : 0,
                        "optime" : Timestamp(0, 0),
                        "optimeDate" : ISODate("1970-01-01T00:00:00Z"),
                        "lastHeartbeat" : ISODate("1970-01-01T00:00:00Z"),
                        "pingMs" : 0,
                        "errmsg" : "socket exception [CONNECT_ERROR] for 172.25.122.186:27017"
                }
        ],
        "ok" : 1
}
monit:SECONDARY>

Executed but still shwoing Secondary and not coming up to Primary

[Asya Kamsky]

It would normally not be instant - what do the logs say?

[Avise Sudhakar Rao]

It was still in Secondary, so I thought as i don't have any other opportunity, created new replicatSet with the name monit2

and after creating new replicatSet 172.25.122.182 is Primary but 172.25.122.186 is becoming to Recovery state and not coming up to Secondary state and the lag on the secondary is keep on increasing.

How can I change oplogsize on primary without restarting Primary? I think if I increase oplogsize of the Primary then this problem will be resolved.

Please help.

[Asya Kamsky]

Yes, it was still in secondary and until you run the rs.reconfig() it will stay in secondary.

When you reconfigure it will be able to become Primary as long as the other nodes can connect to it over the network (if you don't have network connectivity between that node and other nodes, you won't be able to get the replica set back up).

Asya

[Avise Sudhakar Rao]

How to increase oplogsize of primary without stopping it?

monit2:PRIMARY> db.oplog.rs.find().count()
768253
monit2:PRIMARY>

As slave  is not coming up to secondary and it's remaining in recovery state only.

Thanks,
Sudhakar

[Asya Kamsky]

You cannot increase the size of the oplog without stopping the primary.

However, the fastest way to sync a secondary is to snapshot (make a copy) the entire data directory from the primary and drop it into the secondary's data file (when the secondary is stopped).  Then restart the secondary.   It will finish syncing from the oplog of the primary without needing to do a full sync.   Assuming you are running with journaling this does not require stopping the primary while you snapshot the disk.

Asya

[Avise Sudhakar Rao]

If I'm doing this process taking of primary snapshot and placing on slave mongo server when I stop slave mongo server which is in  Recovery State, Primary mongo state getting changed to Secondary and there is no Primary mongo state.

And after copying all the data in the slave mongo db and making the mongo up then this slave mongo state is coming to Recovery and Secondary is automatically getting changed to Primary.

But no way Recovery state is changing to Secondary.

Please suggest how to resolve this.

Thanks,
Sudhakar