We are excited to announce that Elasticsearch 6.3.2 is now available for cluster provisioning on Qbox.io. This is a continuation of our efforts to integrate the latest Elasticsearch versions into Qbox offerings. (We previously made Elasticsearch 6.2.1 clusters available on our platform on March 28, 2018.)
Elasticsearch 6.3.2 is not a major version change as compared to 6.0.0 or 6.2.1, although the release does include several breaking changes we have summarized below. As always, we highly recommend testing in a staging environment before upgrading your Qbox hosted production environment to Elasticsearch 6.3.2. Feel free to contact our 24/7 support team to assist in your migration.
All credit for this minor upgrade goes to the Elasticsearch contributor community under the leadership of Elastic, the company. Much of the documentation cited below links to them.
ElasticHQ was chosen as the default built-in monitoring solution for Qbox-hosted Elasticsearch 6.2.1 and Elasticsearch 6.3.2. The benefit of this tool is that it provides monitoring in real time, is viewable across multiple devices, and has a friendly, user-centric interface (see the image below).
You can check out our latest tutorial about using ElasticHQ to monitor your Qbox-hosted Elasticsearch clusters.
The Qbox.io team is constantly expanding the scope of plugins you can use in Qbox-hosted Elasticsearch clusters. Plugins currently available for the Qbox-hosted Elasticsearch 6.3.2 include the following:
- ICU Analysis
- Japanese (Kuromoji) Analysis
- Smart Chinese Analysis
- Stempel (Polish) Analysis
- Phonetic Analysis
- Ingest Attachment Processor
- Ingest Geo IP Processor
- Ingest User-Agent Processor
To learn more about Elasticsearch community plugins available at Qbox, see our Elasticsearch plugin series.
Improvements found in Elasticsearch 6.x
- Cross-Cluster Search. With the cross-cluster search feature, you can connect a cross-cluster search node to the remote cluster in a light fashion to execute federate search requests without actually joining the remote cluster.
- Quicker Restarts and Recovery. Elasticsearch 6.0.0 introduced sequence IDs that allow operations-based shard recovery by individual shards instead of their resyncing with the primary shard. This approach makes node and shard recovery much faster.
- Sparsely Populated Fields Enhancement to help minimize index storage size. Sparse fields not only require fewer storage slots, but this enhancement also reduces merge times and improves query throughput with no special configuration changes or re-architectures.
- Improved Query Times. With a new index sorting feature, users can now optimize Elasticsearch indexes to store their documents on disk in a specific order. Sorting documents at index time instead of search time dramatically improves query performance and speed.
- Searching Upgrade includes adding scalability to searches across many shards.
- Searching and Indexing New Features includes new aggregations and field types.
Please refer to Elastic’s official Elasicsearch 6.x release page for a complete listing of all improvements.
Many changes have been implemented in Elasticsearch 6.x, so please read over all of Elastic’s official breaking changes documentation to further familiarize yourself with what you may need to change in your code. You can also check out our release highlight for Elasticsearch 6.2.1 for a concise description of Elasticsearch 6.x breaking changes.
Summary of Elasticsearch 6.3 Breaking Changes and New Features
- Suggest stats metrics deprecation. In 6.3.0, the
suggestmetric is deprecated in favor of using
- Plugin changes. A number of changes affected GCS Repository Plugin, Ingest Geoip Plugin, and passing security permissions to Elasticsearch plugin installer. You can find more information here.
- Setting changes including the addition of a small thread pool for analyze operations to make them independent of other threads. You can find more information here.
- Java 10 Support. Elasticsearch added support for Java 10 short-term release. However, the recommended version of Java to use with Elasticsearch 6.3 is still Java 8.
- New Features include Painless execute API, client feature tracking, adding indices options and Search API options support for Rank Evaluation API, introducing new search flags etc. For a full list of new features consult the official Breaking Changes page here.
For further details on all breaking changes implemented in Elasticsearch 6.3 and how they can affect you, please consult Elastic’s official documentation.
When you are ready to migrate your current Qbox Elasticsearch cluster to 6.3.2, please contact the 24/7 support team so that our Support Engineers can assist.
|Production issues||1hr||1 (512) 772-2535|
|Product questions||4hr||Open a ticket|
|Upgrades, maintenance, migrations||8hr||Open a ticket|
Open a ticket
Creating a New Elasticsearch 6.3.2 Cluster on Qbox
To create a new Elasticsearch 6.3.2 cluster in your hosted Qbox environment, please follow these steps:
- Log into your Qbox account
- Click ‘New Cluster’
- Choose ‘AWS’
- Name your cluster
- Select 6.3.2 in the ‘Version’ drop-down
- Determine how much ‘RAM’ you would like
- Decide on how many ‘Processors’ you need
- Assign the number of needed ‘Replicas’
- Click ‘Create’
You can also consult our interactive video tutorial on Spinning Up your Qbox Cluster.