add segment size metric on segment push (#8387)

jadami10 · web-flow · commit 6d2f74978e9f · 2022-03-23T09:50:30.000-07:00
Adds an offline segment size metric only when a new segment is pushed. this is stored as a gauge and based on table name. This is useful to build detectors in case:
- someone tries to push a segment that is too large
- your segments are organically growing in size and you want to know once you've hit a threshold
diff --git a/pinot-common/src/main/java/org/apache/pinot/common/metrics/ControllerGauge.java b/pinot-common/src/main/java/org/apache/pinot/common/metrics/ControllerGauge.java
@@ -69,6 +69,10 @@ public enum ControllerGauge implements AbstractMetrics.Gauge {
   // Estimated size of offline table
   OFFLINE_TABLE_ESTIMATED_SIZE("OfflineTableEstimatedSize", false),
 
+  // Size of the last uploaded offline segment
+  LAST_PUSHED_SEGMENT_SIZE("LastPushedSegmentSize", false),
+
+
   // Table quota based on setting in table config
   TABLE_QUOTA("TableQuotaBasedOnTableConfig", false),
 
diff --git a/pinot-controller/src/main/java/org/apache/pinot/controller/api/resources/PinotSegmentUploadDownloadRestletResource.java b/pinot-controller/src/main/java/org/apache/pinot/controller/api/resources/PinotSegmentUploadDownloadRestletResource.java
@@ -61,6 +61,7 @@
 import org.apache.commons.io.IOUtils;
 import org.apache.commons.lang3.tuple.ImmutablePair;
 import org.apache.commons.lang3.tuple.Pair;
+import org.apache.pinot.common.metrics.ControllerGauge;
 import org.apache.pinot.common.metrics.ControllerMeter;
 import org.apache.pinot.common.metrics.ControllerMetrics;
 import org.apache.pinot.common.restlet.resources.StartReplaceSegmentsRequest;
@@ -304,6 +305,10 @@ private SuccessResponse uploadSegment(@Nullable String tableName, TableType tabl
       completeZkOperations(enableParallelPushProtection, headers, finalSegmentFile, tableNameWithType, segmentMetadata,
           segmentName, zkDownloadUri, moveSegmentToFinalLocation, crypterClassName, allowRefresh);
 
+      // We only set this gauge after the segment is successfully pushed.
+      _controllerMetrics.setValueOfTableGauge(tableNameWithType, ControllerGauge.LAST_PUSHED_SEGMENT_SIZE,
+          FileUtils.sizeOfDirectory(tempSegmentDir));
+
       return new SuccessResponse("Successfully uploaded segment: " + segmentName + " of table: " + tableNameWithType);
     } catch (WebApplicationException e) {
       throw e;
