flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From aljoscha <...@git.apache.org>
Subject [GitHub] flink pull request #2277: [FLINK-4207] WindowOperator becomes very slow with...
Date Mon, 25 Jul 2016 13:55:58 GMT
Github user aljoscha commented on a diff in the pull request:

    --- Diff: flink-streaming-java/src/main/java/org/apache/flink/streaming/runtime/operators/windowing/WindowOperator.java
    @@ -408,11 +410,22 @@ public void processWatermark(Watermark mark) throws Exception {
     				if (windowAssigner instanceof MergingWindowAssigner) {
     					mergingWindows = getMergingWindowSet();
     					W stateWindow = mergingWindows.getStateWindow(context.window);
    +					if (stateWindow == null) {
    +						// then the window is already purged and this is a cleanup
    +						// timer set due to allowed lateness that has nothing to clean,
    +						// so it is safe to just ignore
    +						continue;
    +					}
     					windowState = getPartitionedState(stateWindow, windowSerializer, windowStateDescriptor);
     				} else {
     					windowState = getPartitionedState(context.window, windowSerializer, windowStateDescriptor);
    +				if (windowState.get() == null) {
    --- End diff --
    This doubles the amount of accesses to the state. For example, for RocksDB that's two
DB lookups instead of one. I think we can get the state contents once and pass that to `fireOrContinue()`
instead of the state object. 
    Same for other instances of this snipped. 

If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.

View raw message