Shulou(Shulou.com)06/02 Report--

This article mainly explains "the tuning method of ElasticSearch cluster". The content of the explanation is simple and clear, and it is easy to learn and understand. Please follow the editor's train of thought to study and learn "the tuning method of ElasticSearch cluster".

Delegate power to things that others can do, and focus your free time on lucrative activities. Let go of other useless things and use your time to position yourself.

ElasticSearch cluster tuning

1. Node discovery

Unicast: you can specify the address of the connected host, which is generally set to the address of the master node. If there are multiple es clusters in the cluster environment, this method is generally used for security.

Add the following configuration to elasticSearch.yml

Discovery.zen.ping.unicast.hosts: ["* .61 9301", "* .62: 9301", "10.162.24.63 9301"]

Multicast: the default mode is relatively simple. If there are hosts in the cluster that can receive and respond, you can join the cluster.

Determine whether the host supports multicast:

Use the ifconfig command to query whether the response command contains the MULTICAST attribute.

2. Selection and configuration of nodes

Add the following configuration to elasticSearch.yml

Master node configuration

Node.master: true

Node.data: false

Data node configuration

Node.master: false

Node.data: true

Brain fissure prevention configuration:

Discovery.zen.minimum_master_nodes: n (n: the minimum number of primary nodes qualified and connected to each other, usually 50% of the number of all clusters + 1)

3. Gate of time

Configure safe havens for cluster data and metadata:

Add the following configuration to elasticSearch.yml

Gateway.type=local

4. Configure high query and high throughput

Filter cach

Indices.cache.filter.size= specific memory size or percentage of total memory

Field data cache and circuit breaker

Indices.fielddata.cache.size: 20% (you can set a specific value of 2G, or 20% of total memory) Note: the size of the field data cache is unlimited, so carelessness can cause the cluster memory to explode.

Indices.fielddata.breaker.limit:80% (default is 80%) can also be implemented through a field circuit breaker, setting the memory needed to limit the field cache, and throwing an exception beyond the set value

5. Index buffering and refresh rate

Indices.memory.index_buffer_size: 30% allows you to set the maximum amount of memory for the index, and set 30% of the maximum memory provided for buffering the index. You can also use the indices.memory.min_shard_index_buffer_size property, the default 4mb, to allow a minimum index buffer designed for each shard. The more memory available for the index buffer, the more documents are saved in memory.

Index.refresh_interval attribute. The default refresh rate is 1s (1 second). The refresh rate specifies how fast the document can be visible to the search operation. The equation is very simple: the faster the refresh rate, the slower the query, and the lower the index throughput. If we allow a slower refresh rate, such as 10s or 30s, it is good to set it, which makes ElasticSearch under pressure to update, you can use more resources for indexing and query.

6. Thread pool configuration

Thread_pool.write.queue_size: 2000, configuration write request queue is listed as 2000

It is strongly recommended to adjust the default thread pool and support index queues, write queues, read queues, etc.

7. Template

You can set the common properties of multiple indexes, and the template can be overwritten in layers before, and overwritten sequentially according to the order attribute.

Thank you for your reading. the above is the content of "tuning method of ElasticSearch Cluster". After the study of this article, I believe you have a deeper understanding of the tuning method of ElasticSearch cluster, and the specific use needs to be verified by practice. Here is, the editor will push for you more related knowledge points of the article, welcome to follow!

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.