Migrating to 7.13

edit

This section discusses the changes that you need to be aware of when migrating your application to Elasticsearch 7.13.

See also What’s new in 7.16 and Release notes.

Breaking changes

edit

The following changes in Elasticsearch 7.13 might affect your applications and prevent them from operating normally. Before upgrading to 7.13, review these changes and take the described steps to mitigate the impact.

Breaking changes introduced in minor versions are normally limited to security and bug fixes. Significant changes in behavior are deprecated in a minor release and the old behavior is supported until the next major release. To find out if you are using any deprecated functionality, enable deprecation logging.

Mapping changes

edit
Geo mappers no longer accept external values from multifields.

Details
In earlier versions, geo fields could theoretically be used as a subfield if their parent field set a geometry object as an external value during parsing. This functionality was unused within the Elasticsearch codebase and blocked required changes to allow index-time scripts on geo_point fields, so it has been removed.

Geopoint mappers pass geohashes to subfields one-by-one.

Details
In earlier versions, a multiply-valued geopoint field would pass its values as geohashes to subfields by combining them in to a single comma-delimited string. These are now passed one-by-one as single geohash strings.

SSL/TLS changes

edit
TLSv1.1 and TLSv1.0 are disabled in the bundled JDK

Details
When using the bundled JDK, TLSv1.1 and TLSv1.0 are disabled by default. This may affect SSL connections to the Rest API for some older clients. It also has the potential to affect outgoing connections such as Watcher webhooks, LDAP authentication or access to snapshot repositories.

Most Elasticsearch deployments will not be affected by this change, as these older TLS versions have known vulnerabilities and are no longer heavily used.

For instructions on how to enable these older TLS versions in your Elasticsearch cluster, see Enabling additional SSL/TLS versions on your JDK.

Settings changes

edit
xpack.searchable.snapshot.shared_cache.size is no longer a supported user setting for Elasticsearch Service

Details
You can no longer configure xpack.searchable.snapshot.shared_cache.size on Elasticsearch Service deployments running Elasticsearch 7.13 or a later version. This setting reserves disk space for the shared cache of partially mounted indices. Elasticsearch now automatically configures the setting to 90% of total disk space for frozen data tier nodes and to 0b for non-frozen data tier nodes.

Impact
If you use Elasticsearch Service and previously configured xpack.searchable.snapshot.shared_cache.size, remove it from your user settings before upgrading to 7.13 or a later version. Otherwise, attempts to upgrade the deployment will fail and return an error.

Changes to the frozen tier and multiple data paths

Details
Elasticsearch 7.12 included a technical preview of the frozen tier, being able to use partially mounted indices (searchable snapshots mounted with the shared cache option). Trying out this feature required configuring a shared cache using the xpack.searchable.snapshot.shared_cache.size setting.

In Elasticsearch 7.13+, having a non-zero xpack.searchable.snapshot.shared_cache.size on nodes utilizing multiple data paths (path.data points to multiple locations) is no longer supported and will prevent the node from starting. If you do not utilize multiple data paths this will not affect you. Likewise, if you have not set xpack.searchable.snapshot.shared_cache.size and have not configured dedicated frozen nodes (nodes with the data_frozen role and no other data roles) this will not affect you.

Deprecations

edit

The following functionality has been deprecated in Elasticsearch 7.13 and will be removed in 8.0. While this won’t have an immediate impact on your applications, we strongly encourage you take the described steps to update your code after upgrading to 7.13.

Significant changes in behavior are deprecated in a minor release and the old behavior is supported until the next major release. To find out if you are using any deprecated functionality, enable deprecation logging.

Aggregation deprecations

edit
Date aggregations on boolean fields are deprecated.

Details
Support for auto-interval date histogram, date histogram, and date range aggregations on boolean fields is now deprecated. On boolean fields, these aggregations are rarely useful and often unintended.

Core deprecations

edit
Multiple data path support is deprecated.

Details
The path.data setting accepts a list of data paths, but if you specify multiple paths then the behaviour is unintuitive and usually does not give the desired outcomes. Support for multiple data paths is now deprecated and will be removed in a future release.

Impact
To avoid deprecation warnings, specify a single path in path.data. If needed, you can create a filesystem which spans multiple disks with a hardware virtualisation layer such as RAID, or a software virtualisation layer such as Logical Volume Manager (LVM) on Linux or Storage Spaces on Windows. If you wish to use multiple data paths on a single machine then you must run one node for each data path.

If you currently use multiple data paths in a highly available cluster then you can migrate to a setup that uses a single path for each node without downtime using a process similar to a rolling restart: shut each node down in turn and replace it with one or more nodes each configured to use a single data path. In more detail, for each node that currently has multiple data paths you should follow the following process.

  1. Take a snapshot to protect your data in case of disaster.
  2. Optionally, migrate the data away from the target node by using an allocation filter:

    PUT _cluster/settings
    {
      "persistent": {
        "cluster.routing.allocation.exclude._name": "target-node-name"
      }
    }

    You can use the cat allocation API to track progress of this data migration. If some shards do not migrate then the cluster allocation explain API will help you to determine why.

  3. Follow the steps in the rolling restart process up to and including shutting the target node down.
  4. Ensure your cluster health is yellow or green, so that there is a copy of every shard assigned to at least one of the other nodes in your cluster.
  5. If applicable, remove the allocation filter applied in the earlier step.

    PUT _cluster/settings
    {
      "persistent": {
        "cluster.routing.allocation.exclude._name": null
      }
    }
  6. Discard the data held by the stopped node by deleting the contents of its data paths.
  7. Reconfigure your storage. For instance, combine your disks into a single filesystem using LVM or Storage Spaces. Ensure that your reconfigured storage has sufficient space for the data that it will hold.
  8. Reconfigure your node by adjusting the path.data setting in its elasticsearch.yml file. If needed, install more nodes each with their own path.data setting pointing at a separate data path.
  9. Start the new nodes and follow the rest of the rolling restart process for them.
  10. Ensure your cluster health is green, so that every shard has been assigned.

You can alternatively add some number of single-data-path nodes to your cluster, migrate all your data over to these new nodes using allocation filters, and then remove the old nodes from the cluster. This approach will temporarily double the size of your cluster so it will only work if you have the capacity to expand your cluster like this.

If you currently use multiple data paths but your cluster is not highly available then the you can migrate to a non-deprecated configuration by taking a snapshot, creating a new cluster with the desired configuration and restoring the snapshot into it.

The action.destructive_requires_name setting will default to true in 8.0.0.

Details
In 8.0.0, the action.destructive_requires_name setting will default to true. Currently, the setting defaults to false.

Impact
If you use a wildcard (*) or _all to delete indices or perform other destructive actions, use the update cluster settings API to set action.destructive_requires_name to false to avoid errors in 8.0.0.

index.indexing.slowlog.level and index.search.slowlog.level are deprecated.

Details
The index.indexing.slowlog.level and index.search.slowlog.level index settings are now deprecated. You use these setting to set the logging level for the search and indexing slow logs. To reproduce similar results, use the respective index.*.slowlog.threshold.index.debug and index.*.slowlog.threshold.index.trace index settings instead.

For example, to reproduce a index.indexing.slowlog.level setting of INFO, set index.indexing.slowlog.threshold.index.debug and index.indexing.slowlog.threshold.index.trace to -1.

Impact
To avoid deprecation warnings, discontinue use of the deprecated settings.

EQL deprecations

edit
The wildcard function is deprecated.

Impact
Use the like or regex keyword instead.

Security deprecations

edit
Implicit enablement of the file and native realms is deprecated.

Details
Currently, the file and native realms have the following implicit behaviors:

  • If file and native realms are not configured, they are implicitly disabled if there are other explicitly configured realms.
  • If no realm is available because realms are unconfigured, explicitly disabled, or not allowed by your license, the file and native realms are always enabled, even if explicitly disabled.

Impact
Both of the above behaviors are deprecated. In 8.0.0, the file and native realms will always be enabled unless explicitly disabled. If they are explicitly disabled, they remain disabled at all times.

System call filter setting deprecated

Details
Elasticsearch uses system call filters to remove its ability to fork another process. This is useful to mitigate remote code exploits. These system call filters are enabled by default, and controlled via the setting bootstrap.system_call_filter. Starting in Elasticsearch 8.0, system call filters will be required. As such, the setting bootstrap.system_call_filter is deprecated and will be removed in Elasticsearch 8.0.

Impact
Discontinue use of the removed setting. Specifying this setting in Elasticsearch configuration will result in an error on startup.

Settings deprecations

edit
Several tier filtering settings are deprecated.

Details
The following cluster settings are now deprecated:

  • cluster.routing.allocation.include._tier
  • cluster.routing.allocation.exclude._tier
  • cluster.routing.allocation.require._tier

The following index settings are also now deprecated:

  • index.routing.allocation.include._tier
  • index.routing.allocation.exclude._tier
  • index.routing.allocation.require._tier

These settings are used to filter the allocation of a shard to a specific set of nodes. Instead, use the index.routing.allocation.include._tier_preference index setting.

Impact
To avoid deprecation warnings, discontinue use of the deprecated settings.

The path.shared_data and index.data_path settings are deprecated.

Details
The path.shared_data node setting and index.data_path index setting are now deprecated. Elasticsearch previously used these settings for shadow replicas. The shadow replicas feature was deprecated in 5.2 and removed in 6.0.

Impact
To avoid deprecation warnings, discontinue use of the deprecated settings.