Logstash 6.2.4发布，开源服务端数据处理流程

Logstash 6.2.4 已发布，带来了大量修复和改进：

• Fixed an issue where events passing through the persistent queue had the contents of string type fields corrupted for values containing non-ASCII characters. The solution involved serializing the RubyString as Java String UTF-16 encoded #9167

• Fixed serialization bug when a RubyString that comes out of a matching, substring or similar operation may not have offset 0 pointing at the underlying BytesList. Solved by serializing the correct part of the BytesList #9308

• Improved performance of Event#cancel, where each operation would unnecessarily generate a new object. In configurations that use plugins like the drop filter throughput may increase up to 5x #9284

• Fixed an issue with type handling in metadata fields. Before this a plugin could trigger a Java exception when placing data into an event’s metadata #9299

• Fixed a race condition in shutdown of pipelines where not all workers would consume a single SHUTDOWN signal #9285

• Multiple documentation improvements relating to configuration files, JVM options, default plugin codecs, Logstash-to-Logstash communication and Keystore.

Plugins

Json_lines Codec

• Support flush method, see https://github.com/logstash-plugins/logstash-codec-json_lines/pull/35

Netflow Codec

• Workaround for breaking change in Netflow-Input-UDP > 3.2.0, see issue #122

• Renamed some unknown VMware VDS fields

Aggregate Filter

• new feature: add timeout_timestamp_field option. When set, this option computes timeout based on event timestamp field (and not system time). It’s particularly useful when processing old logs.

• new feature: add inactivity_timeout option. Events for a given task_id will be aggregated for as long as they keep arriving within the defined inactivity_timeout option - the inactivity timeout is reset each time a new event happens. On the contrary, timeout is never reset and happens after timeout seconds since aggregation map creation.

Dns Filter

• Logging improvement to include DNS resolution failure reason #36

• Fix bug where forward lookups would not cache timeout errors

Jdbc_streaming Filter

• Load the driver with the system class loader. Fixes issue loading some JDBC drivers in Logstash 6.2+ #263

Kv Filter

• Correctly handle empty values between value separator and field separator #58

Ruby Filter

• Fix return of multiple events when using file based scripts #41

Translate Filter

• Add refresh_behaviour to either merge or replace during a refresh #57

Beats Input

• Ensure that the keep-alive is sent for ALL pending batches when the pipeline is blocked, not only the batches attempting to write to the queue. #310

Exec Input

• Add metadata data to the event wrt execution duration and exit status

• Add schedule option to schedule the command to run, using a cron expression

Http Input

• Make sure default codec is also cloned for thread safety. #80

• Always flush codec after each request and codec decoding. #81

Jdbc Input

• Clarify use of use_column_value. Make last_run_metadata_path reference in record_last_run entry clickable. #273

• Load the driver with the system class loader. Fixes issue loading some JDBC drivers in Logstash 6.2+ #263

• Fix regression with 4.3.5 that can result in NULL :sql_last_value depending on timestamp format #274

Redis Input

• Add support for SSL #61

• Add support for Redis unix sockets #64

S3 Input

• Improve error handling when listing/downloading from S3 #144

• Add documentation for endpoint, role_arn and role_session_name #142

• Add support for additional_settings option #141

Sqs Input

• Add documentation for endpoint, role_arn and role_session_name #46

• Fix sample IAM policy to match to match the documentation #32

Tcp Input

• Restore SSLSUBJECT field when ssl_verify is enabled. #115

• Update Netty/tc-native versions to match those in beats input #113

Udp Input

• Add metrics support for events, operations, connections and errors produced during execution. #34

• Fix support for IPv6 #31

Aws Mixin

• Drop strict value validation for region option #36

• Add endpoint option to customize the endpoint uri #32

• Allow user to provide a role to assume #27

• Update aws-sdk dependency to ~> 2

Elasticsearch Output

• Set number_of_shards to 1 and document_type to _doc for es 7.x clusters #741 #747

• Fix usage of upsert and script when update action is interpolated #239

• Add metrics to track bulk level and document level responses #585

File Output

• Add feature write_behavior to the documentation #58

S3 Output

• Add documentation for endpoint, role_arn and role_session_name #174

• Add option for additional settings #173

• Add more S3 bucket ACLs #158

• Handle file not found exception on S3 upload #144

• Document prefix interpolation #154

Sqs Output

• Add documentation for endpoint, role_arn and role_session_name #29

Logstash 是一个应用程序日志、事件的传输、处理、管理和搜索的平台。你可以用它来统一对应用程序日志进行收集管理，提供 Web 接口用于查询和统计。Logstash 现在也是 ElasticSearch 家族成员之一。

发布说明：https://www.elastic.co/guide/en/logstash/6.2/logstash-6-2-4.html  
下载地址：https://www.elastic.co/downloads/logstash