jiafu1115 commented on code in PR #20913:
URL: https://github.com/apache/kafka/pull/20913#discussion_r3259505787


##########
storage/src/main/java/org/apache/kafka/server/log/remote/storage/RemoteLogManager.java:
##########
@@ -937,6 +946,68 @@ List<EnrichedLogSegment> candidateLogSegments(UnifiedLog 
log, Long fromOffset, L
             return candidateLogSegments;
         }
 
+        private boolean delayCopy(LogConfig logConfig, LogSegment previousSeg, 
long currentTimeMs, long totalLogSize, long cumulativeSize) {
+            if (logConfig == null) {
+                return false;
+            }
+
+            long copyLagMs = logConfig.remoteCopyLagMs();
+            long copyLagBytes = logConfig.remoteCopyLagBytes();
+            if (logger.isTraceEnabled()) {
+                logger.trace("delayCopy check for segment {}: copyLagMs={}, 
copyLagBytes={}, currentTimeMs={}, totalLogSize={}, cumulativeSize={}, 
sizeLagBytes={}",
+                        previousSeg, copyLagMs, copyLagBytes, currentTimeMs, 
totalLogSize, cumulativeSize, totalLogSize - cumulativeSize);
+            }
+
+            if (copyLagMs == 0 || copyLagBytes == 0) {
+                return false;
+            }
+
+            boolean needCheckCopyLagMs =  copyLagMs > 0;
+            boolean needCheckCopyLagBytes = copyLagBytes > 0;
+
+            // When no lag delay is enabled, upload immediately.
+            if (!needCheckCopyLagMs && !needCheckCopyLagBytes) {

Review Comment:
   @kamalcph 
   thanks for your review. I already split the test to new dedicated test class 
and refactor the code according to part of suggestions. All tests passed except 
flaky tests.
   
   For this question. In your code approach, if the effective localRetentionMs 
and localRetentionSize both are -1
   then effectiveCopyLagBytes and effectiveCopyLagMs both are Long.MAX_VALUE
   then the upload won't be done.
   
   in my code approach. it will be upload at once. So that is the only tiny 
behavior difference.
   
   thus we have another tiny style/performance difference about this behavior 
difference.
   my code approach:
   if one effective configure is MAX value (You use max value. I keep using -1 
as the effective value when local retention is unlimited). I don't check the 
configure.
   your code approach:
   if one is MAX value. you still always check it and get the result is false. 
Take time configure as example: Follow check already been checked.
   `         shouldUploadNow = segmentAgeMs >= effectiveCopyLagMs;`
            
   At the same time. many similar logic for judgement also using the 
retentionMs < 0 (keep -1) instead of using Max.
   ```
       private int deleteRetentionMsBreachedSegments() throws IOException {
           long retentionMs = UnifiedLog.localRetentionMs(config(), 
remoteLogEnabledAndRemoteCopyEnabled());
           if (retentionMs < 0) return 0;         
   ```
   
   WDTY?  Thanks
   
   BTW;
   I add this missed test case. 
   
https://github.com/apache/kafka/pull/20913/changes#diff-a1977b0f066b04154cf4471e9c949ada43119517a345ea5924685f6d34cf1bb2R340



-- 
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