Shulou(Shulou.com)06/01 Report--

Watch the new features of MongoDB3.6 in 3 minutes.

Default Bind to Localhost

Since version 3.6, localhost (127.0.0.1) is used by default, and multiple ip are separated by commas: localhost,198.51.100.1

Except for part of the 2.6 RPM installation package that is local, the rest of the version defaults to All interfaces.

Authentication Restrictions

Starting from version 3.6, the new authenticationRestrictions parameter is used to restrict database user connections to the specified IP address. Add the authenticationRestrictions parameter to the following:

Commands Methods

CreateUser db.createUser ()

UpdateUser db.updateUser ()

CreateRole db.createRole ()

UpdateRole db.updateRole ()

Other security enhancements

When using TLS / SSL encryption, the opensslCipherConfig parameter is added to control the OpenSSL password.

If authentication is enabled, getMore can only be issued against cursors that you create.

A convertToCapped operation was added to restore the role.

Change Streams

MongoDB 3.6supports the use of replica sets or shards using Change Streams. Must be replication protocol version 1 and WT storage engine.

Change Streams allows applications to learn about changes in data in real time without increasing the complexity and risk of tail oplog. Applications can use Change Streams to subscribe to all data changes on the collection and respond to them immediately.

You can use the db.collection.watch () method to open the change stream from any 3. 6 series driver. For complete instructions on how to use it, refer to the documentation for your preferred driver.

The version 3.6 feature parameter featureCompatibilityVersion must be enabled to use Change Streams

See https://docs.mongodb.com/master/reference/command/setFeatureCompatibilityVersion/#view-fcv

When the data reaches the majority members in the cluster, Change Streams immediately notifies the client to respond to these changes.

In clusters with permissions enabled, applications can only use Change Streams to access authorized libraries and collections.

Cursor = db.inventory.watch ()

Document = next (cursor)

Lookup Full Document for Update Operations

Full_document looks at the complete rather than incremental version

Cursor = db.inventory.watch (full_document='updateLookup')

Document = next (cursor)

Resume a Change Stream

Resume_token = document.get ("_ id")

Cursor = db.inventory.watch (resume_after=resume_token)

Document = next (cursor)

Causal Consistency

Requires the client to use MongoDB driver version 3.6and the database to open the 3.6feature parameter featureCompatibilityVersion.

Retryable Writes

Retry will only retry once. For retryable writes, the MongoDB driver automatically retries these operations in the event of a network error or a failover of the replica set, during which the replica set does not have a primary.

Restrictions:

1. Only replica sets and shard are available

two。 Database requires WT or in-memory storage engine

3. The client is required to use MongoDB driver version 3.6, and if necessary, the 3.6feature parameter featureCompatibilityVersion is enabled.

4.writeconcern must be configured and i.e {w 0} is not available.

5. Because the retry attempt is made only once, the retry feature can help resolve temporary network errors, but not persistent network errors.

6. The driver waits for serverSelectionTimeoutMS seconds to determine the new primary node before retrying. The retry feature does not handle situations where the failover period exceeds the serverSelectionTimeoutMS.

Note: if the client application is temporarily unable to respond to the localLogicalSessionTimeoutMinutes after issuing the write operation, the write operation may be retried and reapplied when the client application starts to respond (without restarting).

Serverstatus

Add logicalSessionRecordCache item to serverStatus.

JSON Schema

MongoDB 3.6adds the $jsonSchema operator to support document validation using JSON Schema. For more information, see $jsonSchema.

Replica Sets

Deprecate replica set protocol version 0 (pv0).

The replSetResizeOplog command has been added to dynamically resize the oplog of replica set members. Suitable for instances running the WiredTiger storage engine.

The catchUpTakeoverDelayMillis configuration option has been added to specify how long the node waits before initiating the election, which defaults to 30 seconds.

For replica sets that use protocol version 1 (pv1), if the arbitrator finds that a node with the same or higher priority as the candidate is present, they will vote against it in the election.

Add the oplogInitialFindMaxSeconds parameter to adjust how long members of the replica set wait for their find commands during data synchronization. Default 60s

The waitForSecondaryBeforeNoopWriteMS parameter has been added to specify how long the secondary server must wait if the afterClusterTime is greater than the last application time of the oplog. Default 10 milliseconds

Sharded Clusters

A ShardingTaskExecutorPoolMaxConnecting parameter has been added to mongos to control the rate at which mongos adds connections to the mongod instance. The default is 2, which is valid for mongos only

Added orphanCleanupDelaySecs, which determines the minimum delay before the migration block is removed from the source tile.

You can now fragment the config.system.sessions collection in the config database.

Indexes

Indexes can override queries for fields in nested documents.

If the index tracks which field makes it a multi-key, the multi-key index can override queries for non-array keys.

When you create an index, you cannot specify * as the name of the index.

Listdatabase

Db.adminCommand ({listDatabases: 1, nameOnly: true}) adds nameOnly without locking when executing a command, and if not, a library-level lock is requested.

Db.adminCommand ({listDatabases: 1, filter: {"name": / ^ rep/}}) filter filters the databases you want to see and supports regular expressions

The behavior of the validate command and the db.collection.validate () method has been modified so that only the WiredTiger storage engine enforces a checkpoint, flushes all data in memory to disk, and then validates the data on disk.

The .system.profile entry for update and delete contains the entire update/delete document applied to the named collection.

DropDatabase

The dropDatabase command waits for drop to propagate commands for all collections to most replica set members.

For commands running on replica sets and sharding clusters, the response documents include operationTime and $clusterTime.

Read Concern

Add available. For non-sharded clusters, the "local" and "available" behaviors are the same. For sharding clusters, "available" provides greater tolerance for partitions, but orphaned documents may be returned if shards are undergoing block migration.

Summary

Among the many new features of MongoDB3.6, the reasons for the limited time and space can not be fully explained in this article. We can see that the choice of many new features in the storage engine requires the use of WiredTiger storage engine. The storage engine is also supported by the MongoDB3.0 version, and the MongoDB3.2 version is the default storage engine. Among the many new features, what pleases me most is dynamically resizing oplog, which is very convenient for traditional oplog expansion, and the era of downtime resizing oplog is coming to an end. As the writing time is also very short, it is inevitable that there will be some errors or inaccuracies in the article. I urge readers to criticize and correct them.

Like readers can click like to follow, your praise and attention is the greatest encouragement and support for the author to continue to post!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.