bbeaudreault commented on code in PR #4937:
URL: https://github.com/apache/hbase/pull/4937#discussion_r1082940414


##########
hbase-server/src/main/java/org/apache/hadoop/hbase/namequeues/impl/SlowLogQueueService.java:
##########
@@ -151,15 +156,30 @@ public void consumeEventFromDisruptor(NamedQueuePayload 
namedQueuePayload) {
     final String userName = 
rpcCall.getRequestUserName().orElse(StringUtils.EMPTY);
     final String methodDescriptorName =
       methodDescriptor != null ? methodDescriptor.getName() : 
StringUtils.EMPTY;
-    TooSlowLog.SlowLogPayload slowLogPayload = 
TooSlowLog.SlowLogPayload.newBuilder()
+    TooSlowLog.SlowLogPayload.Builder slowLogPayloadBuilder = 
TooSlowLog.SlowLogPayload.newBuilder()
       .setCallDetails(methodDescriptorName + "(" + param.getClass().getName() 
+ ")")
       
.setClientAddress(clientAddress).setMethodName(methodDescriptorName).setMultiGets(numGets)
       .setMultiMutations(numMutations).setMultiServiceCalls(numServiceCalls)
       .setParam(slowLogParams != null ? slowLogParams.getParams() : 
StringUtils.EMPTY)
       .setProcessingTime(processingTime).setQueueTime(qTime)
       .setRegionName(slowLogParams != null ? slowLogParams.getRegionName() : 
StringUtils.EMPTY)
       
.setResponseSize(responseSize).setServerClass(className).setStartTime(startTime).setType(type)
-      .setUserName(userName).build();
+      .setUserName(userName);
+    if (slowLogParams != null) {
+      if (slowLogParams.getScan() != null) {
+        slowLogPayloadBuilder.setScan(slowLogParams.getScan());
+      }
+      if (slowLogParams.getMulti() != null) {
+        slowLogPayloadBuilder.setMulti(slowLogParams.getMulti());
+      }
+      if (slowLogParams.getGet() != null) {
+        slowLogPayloadBuilder.setGet(slowLogParams.getGet());
+      }
+      if (slowLogParams.getMutate() != null) {
+        slowLogPayloadBuilder.setMutate(slowLogParams.getMutate());

Review Comment:
   i do think we should do that, for one from a privacy/memory perspective, and 
for another to be consistent with the case where cell block encoding is 
enabled. in that case, data won't be in the proto anyway. I don't think the 
literal bytes of the request are super useful, and we already have the total 
size of the request.



-- 
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]

Reply via email to