This is an automated email from the ASF dual-hosted git repository.
martijnvisser pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/flink.git
The following commit(s) were added to refs/heads/master by this push:
new 14df79a3e59 [hotfix][doc] Extensible unified Sink uses new metric to
capture outgoing records
14df79a3e59 is described below
commit 14df79a3e59f77ea5ae0ad2de30fc9b4b478ff0b
Author: Jing Ge <gej...@gmail.com>
AuthorDate: Fri May 6 15:10:03 2022 +0200
[hotfix][doc] Extensible unified Sink uses new metric to capture outgoing
records
---
docs/content/release-notes/flink-1.15.md | 16 ++++++++++++++++
1 file changed, 16 insertions(+)
diff --git a/docs/content/release-notes/flink-1.15.md
b/docs/content/release-notes/flink-1.15.md
index d9669cdff67..5edbb808115 100644
--- a/docs/content/release-notes/flink-1.15.md
+++ b/docs/content/release-notes/flink-1.15.md
@@ -406,6 +406,22 @@ updating the client dependency to a version >= 7.14.0 is
required due to interna
The old JDBC connector (indicated by `connector.type=jdbc` in DDL) has been
removed.
If not done already, users need to upgrade to the newer stack (indicated by
`connector=jdbc` in DDL).
+
+#### Extensible unified Sink uses new metric to capture outgoing records
+
+##### [FLINK-26126](https://issues.apache.org/jira/browse/FLINK-26126)
+
+New metrics `numRecordsSend` and `numRecordsSendErrors` have been introduced
for users to monitor the number of
+records sent to the external system. The `numRecordsOut` should be used to
monitor the number of records
+transferred between sink tasks.
+
+Connector developers should pay attention to the usage of these metrics
numRecordsOut,
+numRecordsSend and numRecordsSendErrors while building sink connectors.
+Please refer to the new Kafka Sink for details.
+Additionally, since numRecordsOut now only counts the records sent between
sink tasks
+and numRecordsOutErrors was designed for counting the records sent to the
external system,
+we deprecated numRecordsOutErrors and recommend using numRecordsSendErrors
instead.
+
## Runtime & Coordination
#### Integrate retry strategy for cleanup stage
