Github user davies commented on a diff in the pull request:

    https://github.com/apache/spark/pull/15923#discussion_r93671999
  
    --- Diff: 
core/src/main/scala/org/apache/spark/storage/ShuffleBlockFetcherIterator.scala 
---
    @@ -305,40 +316,84 @@ final class ShuffleBlockFetcherIterator(
        */
       override def next(): (BlockId, InputStream) = {
         numBlocksProcessed += 1
    -    val startFetchWait = System.currentTimeMillis()
    -    currentResult = results.take()
    -    val result = currentResult
    -    val stopFetchWait = System.currentTimeMillis()
    -    shuffleMetrics.incFetchWaitTime(stopFetchWait - startFetchWait)
    -
    -    result match {
    -      case SuccessFetchResult(_, address, size, buf, isNetworkReqDone) =>
    -        if (address != blockManager.blockManagerId) {
    -          shuffleMetrics.incRemoteBytesRead(buf.size)
    -          shuffleMetrics.incRemoteBlocksFetched(1)
    -        }
    -        bytesInFlight -= size
    -        if (isNetworkReqDone) {
    -          reqsInFlight -= 1
    -          logDebug("Number of requests in flight " + reqsInFlight)
    -        }
    -      case _ =>
    -    }
    -    // Send fetch requests up to maxBytesInFlight
    -    fetchUpToMaxBytes()
     
    -    result match {
    -      case FailureFetchResult(blockId, address, e) =>
    -        throwFetchFailedException(blockId, address, e)
    +    var result: FetchResult = null
    +    var input: InputStream = null
    +    // Take the next fetched result and try to decompress it to detect 
data corruption,
    +    // then fetch it one more time if it's corrupt, throw 
FailureFetchResult if the second fetch
    --- End diff --
    
    @fathersson The checksum in TCP is only 16 bits,  it's not strong enough 
for large traffic, usually DFS or other system with heavy TCP traffic will have 
another application level checksum. Adding to @rxin 's point, we did see this 
retry helped in production to work around temporary corrupt. 


---
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 [email protected] or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to