Elastic Cloud Enterprise 3.2.0edit

The following changes are included in this release.


Simplified deployment creation. This allows the creation of a deployment passing only minimal information. The rest of the information is obtained from the template service without involving the client of the request.

Example request:

curl -XPOST "https://{{hostname}}/api/v1/deployments?template_id={{template_id}}" \
-H "Authorization: ApiKey $ECE_API_KEY" \
-H "Content-Type: application/json" \
-d '
  "version": "7.11.0",
  "region": "local-1"

This makes the resource field of the request optional. This also adds two more fields to the request body: region and version that are used as defaults for all the resources (and the template). If a resource is present in the request and a template is also requested, the content of the request prevails for that resource and no merging of any field is attempted from the template.

Endpoint to download CA certificate for a deployment. Customers can now download the certificate authority used to sign the certificates of their deployment.

Add DirectTrustSetting. Customers can now upload CA certificates to be trusted by individual deployments to configure these as remote connections in Elasticsearch.

Enable using different domain for azure storage. This release enables support for the different domain used by blob storage in Azure Govcloud.


Set the region when creating a dedicated monitoring deployment. Pre-select the correct region when creating a dedicated monitoring deployment.

New ECE diagnostics tool. Added information on Zookeeper transactions and nodes to the output of the ECE diagnostics tool.

Allow list kibana setting elasticsearch.maxSockets: integer. At the moment, the Elasticsearch-js client has a default value of Infinity for the maxSockets parameter. Having such value guarantees that we are not hindering the performance of ES queries below their potential. However, the overhead of establishing TLS sockets can cause performance degradations if the number of concurrent connections is too big.

While there is no consensus as to what is the optimal value for this parameter, some lab tests suggest that 1024 could be a good starting point.

The a in moment.format() is unnecessary when we are using 24 clock. Removing the am/pm notice on upcoming community event dates as we’re already using a 24 hour clock.

Changing the confirmation modal to require the deployment’s name, not the word "DELETE". To confirm the deletion of a deployment now you have to type the name of the deployment instead of the word "DELETE". This makes sure you know which deployment you are deleting.

Bug fixesedit

Fix a Zookeper hang in Container Task Service. Fix a bug which could lead to the Container Task Service not recovering after a Zookeeper outage.

Add a Billing Contact without accepting Cloud’s Terms of Service first. Fixed a bug related to adding a Billing Contact without accepting Cloud’s Terms of Service first. It also fixes a bug in which users can’t leave organizations because they haven’t accepted Terms of Service.

Remove script ready lock. Resolved an issue in the orchestration of containers during an ECE upgrade that may prevent the upgrade’s success.

Remove reference to xpack.reporting.queue.pollEnabled. Removed xpack.reporting.queue.pollEnabled from the allowed config settings for changing user settings.

Remove adminconsole cluster state cache. Filtering by Healthy/Unhealthy shards is no longer available.

  • The GET /deployments, GET /deployments/{deployment_id} and POST /deployments/_search responses include a new shards_status field indicating the Elasticsearch cluster health.
  • GET /deployments/{deployment_id} response includes a new field cluster_blocking_issues with a level indicating global or index and accompanying description of cluster blocks (if any).
  • POST /deployments/_search responses will not return detailed blocks information.
  • All three endpoints will return backwards-compatible empty values in the shards_info and blocking_issues fields.


Replace shards visualization with data from _cluster/health. Replaces the Elasticsearch shards pie visualization with shard counts.