Copilot commented on code in PR #18603:
URL: https://github.com/apache/iotdb/pull/18603#discussion_r3987074987
##########
iotdb-core/datanode/src/main/java/org/apache/iotdb/db/consensus/statemachine/dataregion/DataRegionStateMachine.java:
##########
@@ -55,6 +55,13 @@
import java.util.List;
import java.util.function.Supplier;
+/**
+ * Applies replicated write and query fragment operations to one DataRegion
and exposes its snapshot
+ * and region-resource lifecycle to the consensus layer.
+ *
+ * <p>Write-process rejection is retried here to preserve the atomicity
expected by the consensus
+ * apply path; other statuses are delegated to the consensus retry mechanism.
Review Comment:
`needRetryForSpecificCases` also retries
`METADATA_LEASE_FENCED_RETRY_REQUIRED` for non-pipe writes, while this
class-level text describes only write-process rejection and places other
statuses in the surrounding retry path. Please document both selected statuses
and the pipe-generated exception so the retry contract is accurate.
##########
iotdb-core/datanode/src/main/java/org/apache/iotdb/db/schemaengine/SchemaEngine.java:
##########
@@ -73,7 +73,14 @@
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.TimeUnit;
-// manage all the schemaRegion in this dataNode
+/**
+ * Owns the schema regions hosted by this DataNode and coordinates their
loading, recovery, metrics,
+ * schema-resource management, and lifecycle operations.
+ *
+ * <p>The selected schema-engine mode determines whether schema state is
memory-resident or cached.
+ * Schema metrics and shared schema resources must be initialized before
schema regions, and cleared
+ * only after all schema regions have been cleared.
Review Comment:
The lifecycle described here contradicts `clear()`:
`SchemaResourceManager.clearSchemaResource()` is intentionally called before
each schema region is cleared, while `schemaMetricManager.clear()` runs
afterward. Please document that shared schema resources are cleared before
regions and only schema metrics are cleared last.
##########
iotdb-core/datanode/src/main/java/org/apache/iotdb/db/storageengine/StorageEngine.java:
##########
@@ -912,10 +918,10 @@ public boolean runIfAbsent(DataRegionId regionId,
Runnable runnable) {
}
/**
- * run the consumer if the region is present. if the region is absent, do
nothing.
+ * Run the consumer if the region is present. if the region is absent, do
nothing.
Review Comment:
The second sentence still starts with a lowercase `if`, so the grammar
correction is incomplete.
##########
iotdb-core/datanode/src/main/java/org/apache/iotdb/db/service/DataNode.java:
##########
@@ -252,13 +260,20 @@ public static void main(final String[] args) {
}
}
+ /**
+ * Starts the DataNode by preparing local state, synchronizing cluster
configuration, registering
+ * or restarting the node, recovering regions, and starting the remaining
services in dependency
+ * order.
+ *
+ * <p>The first-start and restart paths intentionally perform different
registration and security
+ * checks.
+ */
@Override
protected void start() {
logger.info(DataNodeMiscMessages.STARTING_DATANODE);
boolean isFirstStart;
try {
-
IoTDBDescriptor.getInstance().getMemoryConfig().activateAutoResizingBufferMemoryControl();
- // Check if this DataNode is start for the first time and do other
pre-checks
+ // Check whether this is the first DataNode startup and run the
remaining startup checks.
Review Comment:
This change removes the call to `activateAutoResizingBufferMemoryControl()`,
so normal DataNode startup no longer installs the automatic RPC-buffer memory
controller. That is a functional regression in a comments-only PR; restore the
activation before `prepareDataNode()` (the method applies the configured
`MemoryConfig` controller).
--
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]