Elastic Cloud Enterprise 2.0.1

edit

Version 2.0.1 of Elastic Cloud Enterprise includes new features, improvements and bug fixes.

New features

edit

Elastic Cloud Enterprise now supports the latest versions of the Elastic Stack, including:

  • Elasticsearch and Kibana 6.5.1
  • Elasticsearch and Kibana 5.6.13

For more information about the features in this version of Elasticsearch and Kibana, see:

To learn more about the best practices to use when upgrading, see Upgrade Your Installation and take note of the updated installation prerequisites for Docker in the next section. For offline (air-gapped) installations: After you complete the upgrade, you must also download the APM server images for the Elastic Stack versions that you plan to use.

Updated prerequisites

edit

New recommended Docker version. After internal testing, we now officially support and recommend Docker 18.09.02 as an installation prerequisite for Elastic Cloud Enterprise. To learn more about the installation of Docker for supported Linux distributions, see Preparing your environment. For existing installations, follow the steps in Perform host maintenance when upgrading to Docker 18.09.02.

Previously, we primarily recommended Docker 1.11 for use with Elastic Cloud Enterprise. We no longer recommend this version, but some Linux distributions might work with other versions of Docker. To see which versions of Docker we support, see Linux distributions.

Installation improvements

edit

If the Docker version is lower than 1.11, the installation script now blocks installation.

Snapshot improvements

edit

ECE now consistently applies a default snapshot interval after enabling snapshots on a deployment.

ZooKeeper improvements

edit

ECE now automatically recovers from allocators failing to authenticate after reconnecting to ZooKeeper. ECE also logs additional information when adding nodes to or removing them from a ZooKeeper ensemble.

Docker improvements

edit

The following improvements for Docker are now available:

  • ECE now ensures that there is no GID collision when adding the elastic user to the docker.sock group.
  • Our documentation now provides updated instructions for using Docker 18.03 on Ubuntu and for using Docker on RHEL. To learn more, see Prepare Your Environment.

API improvements

edit

ECE can now successfully move Elasticsearch nodes for the system-owned admin-console-elasticsearch and the logging-and-metrics deployments from failed allocators.

Proxy Improvements

edit
  • Default max connection limit changed from 2000 max connections to 100000.
  • Default client timeout changed from 50 seconds to 30 seconds.
  • Default server timeout changed from 50 seconds to 6 minutes (default is 10 minutes but in practice it is set to 6 minutes).

Bug fixes

edit

The following bug fixes are included:

  • After an Elasticsearch 6.0 to 6.4 upgrade is complete, ECE now restarts watches to work around a bug in Elasticsearch that may cause inconsistencies after migration of the .watches index. See #33167 for more information.
  • Custom plugins that you uploaded now remain available following a restart of Elasticsearch clusters on versions greater than 6.3.
  • Kibana logs are now always correctly labelled in the UI. Previously, some logs were mislabelled as Elasticsearch logs.

Known problems and limitations

edit

When you use OverlayFS with Kernel-LT 4.4.156 and later, there is a known regression that prevents Elastic Cloud Enterprise from completing the installation. To work around this issue, use Kernel 4.4.155 from the kernel-archive.

Release date: November 27, 2018