Upgrade

edit

The upgrade API allows to upgrade one or more indices to the latest Lucene format through an API. The upgrade process converts any segments written with older formats.

The upgrade API in its current form will not help you to migrate indices created in Elasticsearch 1.x to 5.x.

The upgrade API rewrites an index in the latest Lucene format, but it still retains the original data structures that were used when the index was first created. For instance:

  • Doc-values on numeric fields used to use BinaryDocValues, but now use dedicated NumericDocValues.
  • The parent-child feature has been completely rewritten to use a new data structure.
  • Geo-point fields now require doc values and the Lucene index where, previously, they relied on in-memory calculations.

Migrating 1.x indices to 5.x

The only way to prepare an index created in 1.x for use in 5.x is to reindex your data in a cluster running Elasticsearch 2.3.x, which you can do with the new reindex API.

The steps to do this are as follows:

  1. Create a new index (e.g. new_index) with the correct settings and mappings. These can be retrieved from the old index with the get-index API.
  2. Reindex from old_index to new_index with the reindex API.
  3. Retrieve a list of any aliases associated with the old_index using the get-alias API.
  4. Delete the old_index using the delete index API.
  5. Add an alias called old_index to the new_index along with any aliases returned in step 3, using the update aliases API.

In the future, we plan to change the upgrade API to perform a reindex-in- place. In other words, it would reindex data from old_index to .old_index then atomically delete old_index and rename .old_index to old_index.

Start an upgrade

edit
$ curl -XPOST 'http://localhost:9200/twitter/_upgrade'

Upgrading is an I/O intensive operation, and is limited to processing a single shard per node at a time. It also is not allowed to run at the same time as an optimize/force-merge.

This call will block until the upgrade is complete. If the http connection is lost, the request will continue in the background, and any new requests will block until the previous upgrade is complete.

Request Parameters

edit

The upgrade API accepts the following request parameters:

only_ancient_segments

If true, only very old segments (from a previous Lucene major release) will be upgraded. While this will do the minimal work to ensure the next major release of Elasticsearch can read the segments, it’s dangerous because it can leave other very old segments in sub-optimal formats. Defaults to false.

Check upgrade status

edit

Use a GET request to monitor how much of an index is upgraded. This can also be used prior to starting an upgrade to identify which indices you want to upgrade at the same time.

The ancient byte values that are returned indicate total bytes of segments whose version is extremely old (Lucene major version is different from the current version), showing how much upgrading is necessary when you run with only_ancient_segments=true.

curl 'http://localhost:9200/twitter/_upgrade?pretty&human'
{
  "size": "21gb",
  "size_in_bytes": "21000000000",
  "size_to_upgrade": "10gb",
  "size_to_upgrade_in_bytes": "10000000000"
  "size_to_upgrade_ancient": "1gb",
  "size_to_upgrade_ancient_in_bytes": "1000000000"
  "indices": {
    "twitter": {
      "size": "21gb",
      "size_in_bytes": "21000000000",
      "size_to_upgrade": "10gb",
      "size_to_upgrade_in_bytes": "10000000000"
      "size_to_upgrade_ancient": "1gb",
      "size_to_upgrade_ancient_in_bytes": "1000000000"
    }
  }
}

The level of details in the upgrade status command can be controlled by setting level parameter to cluster, index (default) or shard levels. For example, you can run the upgrade status command with level=shard to get detailed upgrade information of each individual shard.