infraio commented on a change in pull request #2813:
URL: https://github.com/apache/hbase/pull/2813#discussion_r549193188
##########
File path:
hbase-client/src/main/java/org/apache/hadoop/hbase/client/RawAsyncTableImpl.java
##########
@@ -218,50 +222,62 @@ private static Result toResult(HBaseRpcController
controller, MutateResponse res
@Override
public CompletableFuture<Result> get(Get get) {
- return timelineConsistentRead(conn.getLocator(), tableName, get,
get.getRow(),
- RegionLocateType.CURRENT, replicaId -> get(get, replicaId),
readRpcTimeoutNs,
- conn.connConf.getPrimaryCallTimeoutNs(), retryTimer,
conn.getConnectionMetrics());
+ return tracedFuture(
Review comment:
For AsyncTable, its get() method is implemented by RawAsyncTable's get()
method. So here will trace two times? But AsyncTable's get() method's cost time
should be very close with RawAsyncTable's get() method. Do we need to trace
them both? And I thought the "trace tool" may be more useful for RPC call. For
local call stacktrace, profiler tool is more useful?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
