Split Indexedit

The split index API allows you to split an existing index into a new index, where each original primary shard is split into two or more primary shards in the new index.

The number of times the index can be split (and the number of shards that each original shard can be split into) is determined by the index.number_of_routing_shards setting. The number of routing shards specifies the hashing space that is used internally to distribute documents across shards with consistent hashing. For instance, a 5 shard index with number_of_routing_shards set to 30 (5 x 2 x 3) could be split by a factor of 2 or 3. In other words, it could be split as follows:

  • 51030 (split by 2, then by 3)
  • 51530 (split by 3, then by 2)
  • 530 (split by 6)

While you can set the index.number_of_routing_shards setting explicitly at index creation time, the default value depends upon the number of primary shards in the original index. The default is designed to allow you to split by factors of 2 up to a maximum of 1024 shards. However, the original number of primary shards must taken into account. For instance, an index created with 5 primary shards could be split into 10, 20, 40, 80, 160, 320, or a maximum of 640 shards (with a single split action or multiple split actions).

If the original index contains one primary shard (or a multi-shard index has been shrunk down to a single primary shard), then the index may by split into an arbitrary number of shards greater than 1. The properties of the default number of routing shards will then apply to the newly split index.

How does splitting work?edit

Splitting works as follows:

  • First, it creates a new target index with the same definition as the source index, but with a larger number of primary shards.
  • Then it hard-links segments from the source index into the target index. (If the file system doesn’t support hard-linking, then all segments are copied into the new index, which is a much more time consuming process.)
  • Once the low level files are created all documents will be hashed again to delete documents that belong to a different shard.
  • Finally, it recovers the target index as though it were a closed index which had just been re-opened.

Why doesn’t Elasticsearch support incremental resharding?edit

Going from N shards to N+1 shards, aka. incremental resharding, is indeed a feature that is supported by many key-value stores. Adding a new shard and pushing new data to this new shard only is not an option: this would likely be an indexing bottleneck, and figuring out which shard a document belongs to given its _id, which is necessary for get, delete and update requests, would become quite complex. This means that we need to rebalance existing data using a different hashing scheme.

The most common way that key-value stores do this efficiently is by using consistent hashing. Consistent hashing only requires 1/N-th of the keys to be relocated when growing the number of shards from N to N+1. However Elasticsearch’s unit of storage, shards, are Lucene indices. Because of their search-oriented data structure, taking a significant portion of a Lucene index, be it only 5% of documents, deleting them and indexing them on another shard typically comes with a much higher cost than with a key-value store. This cost is kept reasonable when growing the number of shards by a multiplicative factor as described in the above section: this allows Elasticsearch to perform the split locally, which in-turn allows to perform the split at the index level rather than reindexing documents that need to move, as well as using hard links for efficient file copying.

In the case of append-only data, it is possible to get more flexibility by creating a new index and pushing new data to it, while adding an alias that covers both the old and the new index for read operations. Assuming that the old and new indices have respectively M and N shards, this has no overhead compared to searching an index that would have M+N shards.

Preparing an index for splittingedit

Create a new index:

PUT my_source_index
{
  "settings": {
    "index.number_of_shards" : 1
  }
}

In order to split an index, the index must be marked as read-only, and have health green.

This can be achieved with the following request:

PUT /my_source_index/_settings
{
  "settings": {
    "index.blocks.write": true 
  }
}

Prevents write operations to this index while still allowing metadata changes like deleting the index.

Splitting an indexedit

To split my_source_index into a new index called my_target_index, issue the following request:

POST my_source_index/_split/my_target_index
{
  "settings": {
    "index.number_of_shards": 2
  }
}

The above request returns immediately once the target index has been added to the cluster state — it doesn’t wait for the split operation to start.

Indices can only be split if they satisfy the following requirements:

  • the target index must not exist
  • The source index must have fewer primary shards than the target index.
  • The number of primary shards in the target index must be a factor of the number of primary shards in the source index.
  • The node handling the split process must have sufficient free disk space to accommodate a second copy of the existing index.

The _split API is similar to the create index API and accepts settings and aliases parameters for the target index:

POST my_source_index/_split/my_target_index
{
  "settings": {
    "index.number_of_shards": 5 
  },
  "aliases": {
    "my_search_indices": {}
  }
}

The number of shards in the target index. This must be a factor of the number of shards in the source index.

Mappings may not be specified in the _split request.

Monitoring the split processedit

The split process can be monitored with the _cat recovery API, or the cluster health API can be used to wait until all primary shards have been allocated by setting the wait_for_status parameter to yellow.

The _split API returns as soon as the target index has been added to the cluster state, before any shards have been allocated. At this point, all shards are in the state unassigned. If, for any reason, the target index can’t be allocated, its primary shard will remain unassigned until it can be allocated on that node.

Once the primary shard is allocated, it moves to state initializing, and the split process begins. When the split operation completes, the shard will become active. At that point, Elasticsearch will try to allocate any replicas and may decide to relocate the primary shard to another node.

Wait For Active Shardsedit

Because the split operation creates a new index to split the shards to, the wait for active shards setting on index creation applies to the split index action as well.