FrankChen021 commented on code in PR #19572:
URL: https://github.com/apache/druid/pull/19572#discussion_r3420716076
##########
server/src/main/java/org/apache/druid/rpc/ServiceClientFactoryImpl.java:
##########
@@ -30,14 +31,25 @@ public class ServiceClientFactoryImpl implements
ServiceClientFactory
{
private final HttpClient httpClient;
private final ScheduledExecutorService connectExec;
+ private final RequestHeaderContextConfig requestHeaderContextConfig;
public ServiceClientFactoryImpl(
final HttpClient httpClient,
final ScheduledExecutorService connectExec
)
+ {
+ this(httpClient, connectExec, new RequestHeaderContextConfig());
Review Comment:
[P2] Thread configured header mapping through this factory path
The two-argument constructor always installs a default
`RequestHeaderContextConfig`, and production code still uses that constructor
in `SeekableStreamIndexTaskClientFactory` for supervisor task clients. With an
operator mapping such as `X-Trace -> traceId` or `X-Tenant-Id -> tenantId`, the
inbound filter captures the configured context values, but these task-client
RPCs forward only the default `X-Druid-Trace-Id -> traceId` mapping. That drops
custom headers, or rewrites `traceId` to the wrong outbound header name, making
propagation inconsistent for Kafka/Kinesis/Rabbit supervisor task calls. Please
thread the injected `RequestHeaderContextConfig` through this factory path
instead of falling back to defaults.
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
