Recently learned ELK, want to take advantage of this school haven't go to work this period of time to integrate it into the distributed project, but look at the book without integration, want to ask next bosses thought about is this: is the query directly to ES, and then deleted by mysql to update, after finished logstash synchronization, then Redis has priority to cache lookup cache

CodePudding user response:

Elasticsearch: search engine, is also the storage of log,
Logstash: log collector, it receives log input, to some preprocessing of log, and then output to Elasticsearch,
Kibana: UI components, data, and display to the user via the Elasticsearch API,
[the above content from the network]

CodePudding user response:

You come over this baidu's really sward, L synchronization database to ES, increase, deletion occurred at the same time call L, and then the data is a plus for ES no Redis cache, again not mysql, considering the problem is that poor may double writing, ES and database can be synchronized using mysql binlog, read the mysql binlog log, then reads the information into the MQ, then program a MQ consumer consumption MQ, every consumer after a message, the message is written to the ES,