Shulou(Shulou.com)06/01 Report--

This article shows you how to deeply understand the multi-node role configuration of Elasticsearch 5.x cluster. The content is concise and easy to understand, which will definitely brighten your eyes. I hope you can get something through the detailed introduction of this article.

1. The question leads to

The ES5.X node type has more ingest node types. For clusters with 3 nodes, 5 nodes, or more nodes, how to configure node roles to optimize system performance?

2. Overview of node roles in ES2.X and previous versions

3. ES5.X node role list

Because several other types of nodes and uses are easy to understand, there are no more than master nodes, data nodes, and routing nodes.

The purpose of Ingest: 1) the Ingest node is the same as other nodes in the cluster, but it can create multiple processor pipes to modify incoming documents. Similar to the most commonly used Logstash filter has been implemented as a processor.

2) Ingest nodes can be used to perform common data transformations and enrichment. The processor is configured as a pipe. When writing, Ingest Node has 20 built-in processors, such as grok,date,gsub, lowercase / uppercase, delete and rename, etc.

3) before the batch request or index operation, the Ingest node intercepts the request and processes the document. An example of such a processor could be a date processor, which is used to parse a date in a field. Another example is the conversion processor, which converts a field value to a target type, such as a string to an integer.

4. There are many types of ES5.X node combination types, how to set them?

The employee Christian_Dahlqvist of Elasticsearch is interpreted as follows: the default configuration of a node is the combination of master node and data node. For small clusters with 3-5 nodes, all nodes are usually allowed to store data and qualify for master nodes. You can send any request to any node, and because all nodes have copies of the cluster state, they know how to route the request.

Usually only larger clusters can begin to separate dedicated master nodes and data nodes. For many user scenarios, routing nodes are not necessarily required at all.

Dedicated orchestration nodes (also known as client nodes or routing nodes) eliminate the request parsing and final phases of aggregation / query from the data nodes and allow them to focus on processing data. The extent to which this is good for the cluster will vary from situation to situation. Usually I would say that routing nodes are more common when queries are heavily used.

5. How to set the role of nodes in ES5.X cluster

For the configuration of the roles of 3 nodes, 5 nodes or more nodes, there is no clear definition on the Elasticsearch official website, domestic and foreign forums and blogs.

In the development practice, my configuration is as follows (for reference and discussion only) Note: the English in the above screenshot is changed to lowercase.

My thoughts are as follows: 1) for the Ingest node, if we have no requirements such as format conversion, type conversion, etc., set it to false directly. 2) 3-5 nodes belong to lightweight clusters, so make sure that the number of master nodes satisfies ((number of nodes / 2) + 1). 3) lightweight cluster, multiple attributes of nodes, such as Master&Data, can be understood by the same node. 4) if further optimized, the 5 nodes can separate Master from Data and cancel the client node.

The above content is how to deeply understand the multi-node role configuration of Elasticsearch 5.x cluster. have you learned the knowledge or skills? If you want to learn more skills or enrich your knowledge reserve, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.