An open-source appender for Java
Your logs.
Ready to search.
Send Logback events straight to Elasticsearch or OpenSearch. Asynchronous delivery, structured fields, and a familiar XML configuration.
{
"@timestamp": "2026-11-02T10:42:08.000+0000",
"message": "Order accepted",
"level": "INFO"
}
01 / The path from log to index
Keep your logging workflow.
Log through Logback. The appender handles asynchronous delivery to your search cluster.
- 01Your Java appLog as you normally do
- 02Logback appenderEnrich & buffer in memory
- 03Bulk APISend on a worker thread
- 04Your search clusterElasticsearch or OpenSearch
02 / Start with two small changes
A dependency. An appender. Your logs.
Add the library and point Logback at your Bulk API endpoint. The example below targets a local cluster.
1 Add the dependency
<dependency>
<groupId>com.agido</groupId>
<artifactId>logback-elasticsearch-appender</artifactId>
<version>4.0.0</version>
</dependency>Using Gradle? See the complete quick start. Your application supplies Logback and SLF4J.
2 Configure logback.xml
<configuration>
<appender name="ELASTIC"
class="com.agido.logback.elasticsearch.ElasticsearchAppender">
<url>http://localhost:9200/_bulk</url>
<index>application-%date{yyyy-MM-dd}</index>
<headers>
<header>
<name>Content-Type</name>
<value>application/json</value>
</header>
</headers>
</appender>
<root level="INFO">
<appender-ref ref="ELASTIC"/>
</root>
</configuration>@timestamp and message are sent automatically. Add fields through Logback patterns and configuration.
Connecting to a secured cluster? Configure basic authentication or AWS SigV4, and check the tested versions for your stack.
03 / Built for real application logs
More context. Less configuration work.
Asynchronous delivery
A worker thread sends batches, keeping network requests off your application's logging thread.
Explore delivery behavior →Structured context
Include MDC values, SLF4J fluent key-value pairs, and fields generated with Logback patterns.
See field examples →Authentication options
Use basic authentication or AWS Signature Version 4, including the service name for OpenSearch Serverless.
Configure authentication →Control the tradeoffs
Set batch size, buffer limits, timeouts and retries to fit your application's delivery requirements.
Browse the reference →An independently maintained fork
Coming from internetitem?
This independently maintained fork brings dependency updates, bug fixes and current platform support. Review the explicit migration steps, including package names and shutdown behavior.
04 / Know what you are running
Open source. Documented limits.
The buffer lives in memory. Outages, exhausted retries or shutdown can leave logs undelivered. For durable, guaranteed delivery, use an external log shipper or queue.
Understand delivery guarantees and limits →