Database Manual / Sharding

Data Partitioning with Chunks基于块的数据分区

MongoDB uses the shard key associated to the collection to partition the data into chunks owned by a specific shard. MongoDB使用与集合关联的分片键将数据划分为特定分片所拥有的块。A chunk consists of a range of sharded data. 块由一系列分片数据组成。A range can be a portion of the chunk or the whole chunk. 范围可以是块的一部分或整个块。The balancer migrates data between shards. Each chunk has inclusive lower and exclusive upper limits based on the shard key.平衡器在分片之间迁移数据。每个块都有基于分片键的包容性下限和排他性上限。

Diagram of the shard key value space segmented into smaller ranges or chunks.

The smallest unit of data a chunk can represent is a single unique shard key value.块可以表示的最小数据单位是单个唯一的分片键值。

Initial Chunks初始块

Populated Collection已填充集合

  • The sharding operation creates one large initial chunk to cover all of the shard key values.分片操作创建了一个大的初始块来覆盖所有分片键值。
  • After the initial chunk creation, the balancer moves ranges off of the initial chunk when it needs to start balancing data.在初始块创建之后，平衡器在需要开始平衡数据时将范围移出初始块。

Empty Collection空集合

  • If you have zones and zone ranges defined for an empty or non-existing collection.如果为空集合或不存在的集合定义了分区和分区范围。

    • The sharding operation creates empty chunks for the defined zone ranges as well as any additional chunks to cover the entire range of the shard key values and performs an initial chunk distribution based on the zone ranges. This initial creation and distribution of chunks allows for faster setup of zoned sharding.分片操作为定义的区域范围创建空块以及任何额外的块，以覆盖分片键值的整个范围，并根据区域范围执行初始块分布。这种块的初始创建和分发允许更快地设置分区分片。
    • After the initial distribution, the balancer manages the chunk distribution going forward.在初始分发之后，平衡器管理接下来的块分发。

  • If you do not have zones and zone ranges defined for an empty or non-existing collection:如果您没有为空或不存在的集合定义区域和区域范围：

    • For hashed sharding:对于哈希分片：

      • The sharding operation creates empty chunks to cover the entire range of the shard key values and performs an initial chunk distribution. By default, the operation creates 2 chunks per shard and migrates across the cluster.分片操作创建空块以覆盖分片键值的整个范围，并执行初始块分布。默认情况下，该操作在每个分片上创建2个块，并在集群中迁移。
      • After the initial distribution, the balancer manages the chunk distribution going forward.在初始分发之后，平衡器管理接下来的块分发。

    • For ranged sharding:对于远程分片：

      • The sharding operation creates a single empty chunk to cover the entire range of the shard key values.分片操作创建了一个空块来覆盖分片键值的整个范围。
      • After the initial chunk creation, the balancer migrates the initial chunk across the shards as appropriate as well as manages the chunk distribution going forward.在初始块创建之后，平衡器会根据需要在分片之间迁移初始块，并管理未来的块分布。

Tip

sh.balancerCollectionStatus()

Range Size分布范围

The default range size in MongoDB is 128 megabytes. You can increase or reduce the chunk size. MongoDB中的默认范围大小为128 MB。您可以增加或减少块大小。Consider the implications of changing the default chunk size:考虑更改默认块大小的影响：

  1. Small ranges lead to a more even distribution of data at the expense of more frequent migrations. This creates expense at the query routing (mongos) layer.较小的范围会导致数据分布更加均匀，但会以更频繁的迁移为代价。这会在查询路由（mongos）层产生开销。
  2. Large ranges lead to fewer migrations. This is more efficient both from the networking perspective and in terms of internal overhead at the query routing layer. But, these efficiencies come at the expense of a potentially uneven distribution of data.大范围导致迁移减少。从网络角度和查询路由层的内部开销来看，这都更有效。但是，这些效率是以潜在的数据不均匀分布为代价的。
  3. Range size affects the Maximum Number of Documents Per Range to Migrate.范围大小会影响每个范围要迁移的最大文档数。

For many deployments, it makes sense to avoid frequent and potentially spurious migrations at the expense of a slightly less evenly distributed data set.对于许多部署，以牺牲分布稍不均匀的数据集为代价避免频繁和潜在的虚假迁移是有意义的。

Range Migration距离迁移

MongoDB migrates data ranges in a sharded cluster to distribute the data of a sharded collection evenly among shards. MongoDB迁移分片集群中的数据范围，以在分片之间均匀分布分片集合的数据。Migrations may be either:迁移可以是：

  • Manual. Only use manual migration in limited cases, such as to distribute data during bulk inserts. 手册。仅在有限的情况下使用手动迁移，例如在批量插入期间分发数据。See Migrating Chunks Manually for more details.有关更多详细信息，请参阅手动迁移块。
  • Automatic. The balancer process automatically migrates data when there is an uneven distribution of a sharded collection's data across the shards. 自动。当分片集合的数据在分片之间分布不均时，平衡器进程会自动迁移数据。See Migration Thresholds for more details.有关更多详细信息，请参阅迁移阈值。

For more information on the sharded cluster balancer, see Sharded Cluster Balancer.有关分片集群平衡器的更多信息，请参阅分片集群均衡器。

Tip

shardingStatistics.countDonorMoveChunkLockTimeout

Balancing平衡

The balancer is a background process that manages data migrations. 平衡器是一个管理数据迁移的后台进程。If the difference in amount of data between the largest and smallest shard exceed the migration thresholds, the balancer begins migrating data across the cluster to ensure an even distribution.如果最大和最小分片之间的数据量差异超过迁移阈值，平衡器将开始在整个集群中迁移数据，以确保均匀分布。

Diagram of a collection distributed across three shards. For this collection, the difference in the number of chunks between the shards reaches the *migration thresholds* (in this case, 2) and triggers migration.

You can manage certain aspects of the balancer. The balancer also respects any zones created as a part of configuring zones in a sharded cluster.您可以管理平衡器的某些方面。平衡器还尊重在分片集群中配置区域时创建的任何区域。

See Sharded Cluster Balancer for more information on the balancer.有关平衡器的更多信息，请参阅分片集群平衡器。

Indivisible/Jumbo Chunks不可分割/大块

In some cases, chunks can grow beyond the specified chunk size but cannot undergo a split. The most common scenario is when a chunk represents a single shard key value. 在某些情况下，块可以增长到指定的块大小之外，但不能进行分割。最常见的情况是，一个块代表一个分片键值。Since the chunk cannot split, it continues to grow beyond the chunk size, becoming a jumbo chunk. 由于块无法分割，它会继续增长到块大小之外，成为一个巨型块。These jumbo chunks can become a performance bottleneck as they continue to grow, especially if the shard key value occurs with high frequency.随着这些巨型块的不断增长，它们可能会成为性能瓶颈，特别是在分片键值频繁出现的情况下。

Starting in MongoDB 5.0, you can reshard a collection by changing a document's shard key.从MongoDB 5.0开始，您可以通过更改文档的分片键来重新分片集合。

MongoDB provides the refineCollectionShardKey command. MongoDB提供了refineCollectionShardKey命令。Refining a collection's shard key allows for a more fine-grained data distribution and can address situations where the existing key insufficient cardinality leads to jumbo chunks.细化集合的分片键可以实现更细粒度的数据分布，并可以解决现有键基数不足导致巨型块的情况。

To learn whether you should reshard your collection or refine your shard key, see Change a Shard Key.要了解是否应该重新标记集合或优化分片键，请参阅更改分片键。

For more information, see:有关更多信息，请参阅：