xintongsong commented on code in PR #22224:
URL: https://github.com/apache/flink/pull/22224#discussion_r1181538748
##########
flink-core/src/main/java/org/apache/flink/configuration/ClusterOptions.java:
##########
@@ -140,7 +140,7 @@ public class ClusterOptions {
public static final ConfigOption<Boolean>
ENABLE_FINE_GRAINED_RESOURCE_MANAGEMENT =
ConfigOptions.key("cluster.fine-grained-resource-management.enabled")
.booleanType()
- .defaultValue(false)
+ .defaultValue(true)
Review Comment:
In addition to changing the default value, I think we should annotate the
configuration as `@Deprecated` and exclude it from documentation. Because we
are expecting to remove the declarative slot manager in the next release, and
this config will then become invalid.
##########
tools/ci/stage.sh:
##########
@@ -26,7 +26,7 @@ STAGE_CONNECTORS_2="connect_2"
STAGE_TESTS="tests"
STAGE_MISC="misc"
STAGE_CLEANUP="cleanup"
-STAGE_FINEGRAINED_RESOURCE_MANAGEMENT="finegrained_resource_management"
+STAGE_DECLARATIVE_RESOURCE_MANAGEMENT="declarative_resource_management"
Review Comment:
I wonder if we can simply remove this stage. I'd expect there to be barely
any changes to the declarative slot manager before it gets removed. Maybe
relying on the unit tests would be good enough.
##########
docs/content/docs/deployment/finegrained_resource.md:
##########
@@ -84,10 +84,12 @@ To use fine-grained resource management, you need to:
### Enable Fine-Grained Resource Management
-To enable fine-grained resource management, you need to configure the
[cluster.fine-grained-resource-management.enabled]({{< ref
"docs/deployment/config" >}}#cluster-fine-grained-resource-management-enabled)
to true.
+The fine-grained resource management is enabled by default.
+
+To ensure this, you need to check that
[cluster.fine-grained-resource-management.enabled]({{< ref
"docs/deployment/config" >}}#cluster-fine-grained-resource-management-enabled)
is not set to false.
{{< hint danger >}}
-Without this configuration, the Flink runtime cannot schedule the slots with
your specified resource requirement and the job will fail with an exception.
+When this configuration is disabled, the Flink runtime cannot schedule the
slots with your specified resource requirement and the job will fail with an
exception.
{{< /hint >}}
Review Comment:
And the *Usage* section should also be updated, because "configure to enable
fine-grained resource management" is no longer needed.
##########
tools/ci/stage.sh:
##########
@@ -26,7 +26,7 @@ STAGE_CONNECTORS_2="connect_2"
STAGE_TESTS="tests"
STAGE_MISC="misc"
STAGE_CLEANUP="cleanup"
-STAGE_FINEGRAINED_RESOURCE_MANAGEMENT="finegrained_resource_management"
+STAGE_DECLARATIVE_RESOURCE_MANAGEMENT="declarative_resource_management"
Review Comment:
Either way, we would need a JIRA ticket to remind us to remove declarative
slot manager in the next release. A ticket with Blocker priority and 1.19/2.0
fixVersion should do the job.
##########
docs/content/docs/deployment/finegrained_resource.md:
##########
@@ -84,10 +84,12 @@ To use fine-grained resource management, you need to:
### Enable Fine-Grained Resource Management
-To enable fine-grained resource management, you need to configure the
[cluster.fine-grained-resource-management.enabled]({{< ref
"docs/deployment/config" >}}#cluster-fine-grained-resource-management-enabled)
to true.
+The fine-grained resource management is enabled by default.
+
+To ensure this, you need to check that
[cluster.fine-grained-resource-management.enabled]({{< ref
"docs/deployment/config" >}}#cluster-fine-grained-resource-management-enabled)
is not set to false.
{{< hint danger >}}
-Without this configuration, the Flink runtime cannot schedule the slots with
your specified resource requirement and the job will fail with an exception.
+When this configuration is disabled, the Flink runtime cannot schedule the
slots with your specified resource requirement and the job will fail with an
exception.
{{< /hint >}}
Review Comment:
I think we can simply remove this section. Users now can use either
coarse-grained or fine-grained resource management, by simply specifying
fine-grained resource requirements in their (DataStream) codes or not, without
having to worry about whether the feature is enabled.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
