guihecheng commented on a change in pull request #3124:
URL: https://github.com/apache/ozone/pull/3124#discussion_r812525131
##########
File path:
hadoop-ozone/client/src/main/java/org/apache/hadoop/ozone/client/io/ECKeyOutputStream.java
##########
@@ -380,42 +380,49 @@ void writeParityCells(int parityCellSize) throws
IOException {
}
private int handleDataWrite(int currIdx, byte[] b, int off, long len,
- boolean isFullCell) throws IOException {
+ boolean isFullCell) {
int pos = ecChunkBufferCache.addToDataBuffer(currIdx, b, off, (int) len);
- handleOutputStreamWrite(currIdx, len, isFullCell, false);
if (pos == ecChunkSize) {
+ handleOutputStreamWrite(currIdx, pos, isFullCell, false);
blockOutputStreamEntryPool.getCurrentStreamEntry().useNextBlockStream();
}
return pos;
}
- private void handleParityWrite(int currIdx, long len, boolean isFullCell) {
+ private void handleParityWrite(int currIdx, int len, boolean isFullCell) {
handleOutputStreamWrite(currIdx, len, isFullCell, true);
blockOutputStreamEntryPool.getCurrentStreamEntry().useNextBlockStream();
}
- private void handleOutputStreamWrite(int currIdx, long len,
+ private void handleOutputStreamWrite(int currIdx, int len,
boolean isFullCell, boolean isParity) {
- BlockOutputStreamEntry current =
+ ECBlockOutputStreamEntry current =
blockOutputStreamEntryPool.getCurrentStreamEntry();
if (isFullCell) {
ByteBuffer bytesToWrite = isParity ?
ecChunkBufferCache.getParityBuffers()[currIdx - numDataBlks] :
ecChunkBufferCache.getDataBuffers()[currIdx];
try {
- // Since it's a fullcell, let's write all content from buffer.
- writeToOutputStream(current, len, bytesToWrite.array(),
- bytesToWrite.limit(), 0, isParity);
+ // Since it's a full cell, let's write all content from buffer.
+ // At a time we write max cell size in EC. So, it should safe to cast
+ // the len to int to use the super class defined write API.
+ // The len cannot be bigger than cell buffer size.
+ assert len <= ecChunkSize : " The len: " + len + ". EC chunk size: "
+ + ecChunkSize;
+ assert len <= bytesToWrite
+ .limit() : " The len: " + len + ". Chunk buffer limit: "
+ + bytesToWrite.limit();
+ writeToOutputStream(current, bytesToWrite.array(), len, 0, isParity);
Review comment:
Hi Uma, So here is the core fix for the problem using len instead of
bytesToWrite.limit() ?
Happen to notice that `handleOutputStreamWrite` only functions when
`isFullCell` and are only called when full cell check is already done. So maybe
there are cleanups to do for this logic ?
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
