In Elastic search doe does slow log level means. There are multiple level, warning,info,trace,debug. If I want to see which querys are taking more than 3 sec which level do I select. what is the difference between trace and debug. which is more verbose

CodePudding user response：

In all these log levels of slow logs, you will see same information like slow log query, index, shard, total_hits etc, main difference is what should be the level for different threshold of slow log time in your application.

For example: if you think query taking beyond 3 seconds is very slow in your application, then you can log it at ERROR level, so that later on you can easily filter them based on different levels(in log analysis tools), but if you think 3 seconds is pretty OK for your application(for some application(not real time analytics application) ES queries taking 3 second might be normal, so they can choose them to log them at INFO or lower levels.

In-short, these logs levels doesn't related to Elasticsearch levels but more related to the application's SLA which are using Elasticsearch, hence every application can customize them, hence Elasticsearch provided this customisation ability.