9aman commented on code in PR #15008:
URL: https://github.com/apache/pinot/pull/15008#discussion_r1955871640


##########
pinot-controller/src/main/java/org/apache/pinot/controller/validation/RealtimeSegmentValidationManager.java:
##########
@@ -135,6 +138,18 @@ private boolean shouldEnsureConsuming(String 
tableNameWithType) {
     if (isTablePaused && 
pauseStatus.getReasonCode().equals(PauseState.ReasonCode.ADMINISTRATIVE)) {
       return false;
     }
+    try {
+      boolean isResourceUtilizationWithinLimits =
+          
_resourceUtilizationManager.isResourceUtilizationWithinLimits(tableNameWithType);
+      if (!isResourceUtilizationWithinLimits) {
+        LOGGER.warn("Resource utilization limit exceeded for table: {}", 
tableNameWithType);
+        _llcRealtimeSegmentManager.pauseConsumption(tableNameWithType,

Review Comment:
   For pauseless this is a bit scary as for the:
   1. Normal Ingestion: Lack of disk will lead to build failure pausing 
ingestion. 
   2. Pausless: New segments will get created till we mark the table as paused 
in IS. Build failure will not block this. 
    Storage issues during the ingestion (before build) when we mmap fetched 
messages will fail. 
    
   Assuming all replicas run out of disk:
   1.  For the segments that only fail during build phase: We need to rely on 
DR to fix these. 
   2. For segments that fail during ingestion (before build) : the rows will be 
marked errored rows and the mutable segment will fail build. This too will be 
fixed using DR.



-- 
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: commits-unsubscr...@pinot.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org
For additional commands, e-mail: commits-h...@pinot.apache.org

Reply via email to