A, JVM configuration:
Jdk1.8 (es requires a minimum jdk1.7)
Maximum heap heap are the smallest 32G
GC type: -XX: + UseG1Gc -XX: MaxGCPauseMillis = 200.
Other default.
Two, Es cluster configuration:
cluster.name: the Test-es
node.name: "the Test-169-71"
node.master: false
node.data: to true
index.store.type: niofs reading and writing papers
index.cache.field .type: soft buffer type
bootstrap.mlockall: true disables the swap
gateway.type: local local storage
gateway.recover_after_nodes: 3 3 data nodes resumed
gateway.recover_after_time: began to recover data 5m 5 minutes
gateway.expected_nodes: 4 4 th es node resumed
cluster.routing.allocation.node_initial_primaries_recoveries: 8 number of concurrent recovery fragments
cluster.routing.allocation.node_concurrent_recoveries: 2 while the number of concurrent recovery
indices.recovery.max_bytes_per_sec: 250mb maximum data transfer bandwidth between nodes
indices.recovery.concurrent_streams: 8 threads simultaneously read data file stream
discovery.zen.ping.multicast.enabled: false to disable multicast
discovery.zen.ping.unicast.hosts: [ "192.168.169.51:9300", "192.168.169.52:9300"]
discovery.zen.fd.ping_interval: 10s between detection intervals surviving nodes
discovery.zen.fd.ping_timeout: 120s survival timeout
discovery.zen.fd.ping_retries: 6 alive timeout retries
http.cors.enabled: true usage monitoring
index.analysis.analyzer.ik.type: "ik" ik word
Setting the Thread the pool ################# #####
threadpool.index.type: Fixed index write a thread pool type
threadpool.index.size: 64 thread pool size (Recommendation 2 to 3 times the number of CPU)
threadpool.index.queue_size: 1000 queue size
threadpool.search.size: 64 search thread pool size
threadpool.search.type: fixed search thread pool type
threadpool.search.queue_size: 1000 queue size
threadpool.get. type: fixed thread pool access to data types
threadpool.get.size: 32 fetch thread pool size
threadpool.get.queue_size: 1000 queue size
threadpool.bulk.type: fixed batch request thread pool type
threadpool.bulk.size: 32 batch request thread pool size
threadpool.bulk.queue_size: 1000 queue size
threadpool.flush.type: fixed disk brush type thread pool
threadpool.flush.size: 32 brush disk thread pool size
threadpool.flush.queue_size: 1000 queue size
# indices.store. throttle.type: merge
indices.store.throttle.type: none write disk type
indices.store.throttle.max_bytes_per_sec: 500mb write disk maximum bandwidth
index.merge.scheduler.max_thread_count: maximum number of threads 8 Index Merge
index.translog.flush_threshold_size: 600MB file, translog refresh valve value