Back Up Sharded Clusters with Database Dumps

Find a Backup Window

To find a good time to perform a backup, monitor your application and database usage to find a time when chunk migrations, resharding, and schema transformation operations are unlikely to occur.

For more information, see Schedule Backup Window for Sharded Clusters.

Stop the Balancer

To prevent chunk migrations from disrupting the backup, connect to mongos and use the sh.stopBalancer() method to stop the balancer:

sh.stopBalancer()

If a balancing round is in progress, the operation waits for balancing to complete.

To verify that the balancer is stopped, use the sh.getBalancerState() method:

use config
while( sh.isBalancerRunning().mode != "off" ) {
   print( "Waiting for Balancer to stop..." );
   sleep( 1000 );
}

Lock the Cluster

The sharded cluster must remain locked during the backup process to protect the database from writes, which may cause inconsistencies in the backup.

To lock a sharded cluster, connect to mongos and use the db.fsyncLock() method:

db.getSiblingDB("admin").fsyncLock()

To confirm the lock, on mongos and the primary mongod of the config servers, run the following aggregation pipeline and ensure that all the shards are locked:

db.getSiblingDB("admin").aggregate( [
   {  $currentOp: { } },
   {  $facet: {
      "locked": [
         { $match: { $and: [
            { fsyncLock: { $exists: true } }
         ] } }],
       "unlocked": [
         { $match: { fsyncLock: { $exists: false } } }
       ]
   } },
   {  $project: {
      "fsyncLocked": { $gt: [ { $size: "$locked" }, 0 ] },
      "fsyncUnlocked": { $gt: [ { $size: "$unlocked" }, 0 ] }
   } }
 ] )

[ { fsyncLocked: true }, { fsyncUnlocked: false } ]

Take Backup

To back up the sharded cluster, use mongodump to connect to mongos and perform the backup:

mongodump \
   --host mongos.example.net \
   --port 27017 \
   --username user \
   --password "passwd" \
   --out /opt/backups/example-cluster-1

Unlock the Cluster

After the backup completes, you can unlock the cluster to allow writes to resume.

To unlock the cluster, use the db.fsyncUnlock() method:

db.getSibling("admin").fsyncUnlock()

To confirm the unlock, on mongos and the primary mongod of the config servers, run the following aggregation pipeline and ensure that all shards are unlocked:

db.getSiblingDB("admin").aggregate( [
   {  $currentOp: { } },
   {  $facet: {
      "locked": [
         { $match: { $and: [
            { fsyncLock: { $exists: true } }
         ] } }],
       "unlocked": [
         { $match: { fsyncLock: { $exists: false } } }
       ]
   } },
   { $project: {
      "fsyncLocked": { $gt: [ { $size: "$locked" }, 0 ] },
      "fsyncUnlocked": { $gt: [ { $size: "$unlocked" }, 0 ] }
   } }
 ] )

[ { fsyncLocked: false }, { fsyncUnlocked: true } ]

Restart the Balancer

To restart the balancer, use the sh.startBalancer() method:

sh.startBalancer()

To confirm that the balancer is running, use the sh.getBalancerState() method:

sh.getBalancerState()

true

The command returns true when the balancer is running.