<Kafka name="Kafka" topic="XX_log"> <PatternLayout pattern="%d{yyyy-MM-dd HH:mm:ss}||%p||%c{1}||XX_web||%m%n"/> <Property name="bootstrap.servers">127.0.0.1:9092</Property> <Property name="timeout.ms">500</Property> </Kafka>
The format in PatternLayout uses || to connect the content for the purpose of splitting logstash. The timeout.ms attribute is added to ensure that the log system hangs up without a great impact on the business system. Of course, kafka can use the cluster method. Multiple addresses of bootstrap.servers are separated by ",". XX_web represents the current business platform.
2. Build a kafka cluster. I won't introduce more here. The official website is very complete.
zookeeper.connect=127.0.0.1:2181,127.0.0.1:2182,127.0.0.1:2183
3. Create a logstash dynamic template
{ "template": "*", "settings": { "index.refresh_interval": "5s", "number_of_replicas": "0", "number_of_shards": "3" }, "mappings": { "_default_": { "_all": { "enabled": false }, "dynamic_templates": [ { "message_field": { "match": "message", "match_mapping_type": "string", "mapping": { "type": "string", "index": "analyzed" } } }, { "string_fields": { "match": "*", "match_mapping_type": "string", "mapping": { "type": "string", "index": "not_analyzed" } } } ], "properties": { "dateTime": { "type": "date", "format": "yyy-MM-dd HH:mm:ss" }, "@version": { "type": "integer", "index": "not_analyzed" }, "context": { "type": "string", "index": "analyzed" }, "level": { "type": "string", "index": "not_analyzed" }, "class": { "type": "string", "index": "not_analyzed" }, "server": { "type": "string", "index": "not_analyzed" } } } } }
4. Configure logstash
input{ kafka { zk_connect =>"127.0.0.1:2181,127.0.0.1:2182,127.0.0.1:2183" group_id =>"logstash" topic_id =>"XX_log" reset_beginning => false consumer_threads => 5 decorate_events => true } } filter { mutate{ split=>["message","||"] add_field => { "dateTime" => "%{[message][0]}" } add_field => { "level" => "%{[message][1]}" } add_field => { "class" => "%{[message][2]}" } add_field => { "server" => "%{[message][3]}" } add_field => { "context" => "%{[message][4]}" } remove_field => ["message"] } date { match => ["logdate", "yyyy-MM-dd HH:mm:ss"] } } output{ elasticsearch { hosts => ["127.0.0.1:9200"] index => "XX_log-%{+YYYY-MM}" codec => "json" manage_template => true template_overwrite => true flush_size => 50000 idle_flush_time => 10 workers => 2 template => "E:\logstash\template\template_log.json" } }
Save the log into the ES index according to the year and month index => "XX_log-%{+YYYY-MM}", logstash reads the log information from the kafka cluster.
5. Building a ZK cluster, I won't introduce it here, there are more online information----http://blog.csdn.net/shirdrn/article/details/7183503
6. Building an ES cluster, the ES cluster is relatively simple to set up The parameters can be used without too many. http://blog.csdn.net/xgjianstart/article/details/52192675
7. Configure kibana
server.port: 5601 # service port # The host to bind the server to. server.host: "115.28.240.113" elasticsearch.url: http://127.0.0.1:9200 ES address-cluster kibana.index: "kibana"
8, Print Book JKD 1.7 ES-2.4, logstash 2.4, kafka-2.10, kibana-4.6.4