sunhelly commented on a change in pull request #2699:
URL: https://github.com/apache/hbase/pull/2699#discussion_r533076760
##########
File path:
hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileInfo.java
##########
@@ -359,24 +360,32 @@ public void initMetaAndIndex(HFile.Reader reader) throws
IOException {
// Initialize an block iterator, and parse load-on-open blocks in the
following.
blockIter = blockReader.blockRange(trailer.getLoadOnOpenDataOffset(),
context.getFileSize() - trailer.getTrailerSize());
- // Data index. We also read statistics about the block index written after
- // the root level.
- this.dataIndexReader = new HFileBlockIndex
- .CellBasedKeyBlockIndexReader(trailer.createComparator(),
trailer.getNumDataIndexLevels());
-
dataIndexReader.readMultiLevelIndexRoot(blockIter.nextBlockWithBlockType(BlockType.ROOT_INDEX),
- trailer.getDataIndexCount());
- reader.setDataBlockIndexReader(dataIndexReader);
- // Meta index.
- this.metaIndexReader = new HFileBlockIndex.ByteArrayKeyBlockIndexReader(1);
-
metaIndexReader.readRootIndex(blockIter.nextBlockWithBlockType(BlockType.ROOT_INDEX),
+ try {
+ // Data index. We also read statistics about the block index written
after
+ // the root level.
+ this.dataIndexReader =
+ new
HFileBlockIndex.CellBasedKeyBlockIndexReader(trailer.createComparator(),
trailer.getNumDataIndexLevels());
+ dataIndexReader
+
.readMultiLevelIndexRoot(blockIter.nextBlockWithBlockType(BlockType.ROOT_INDEX),
trailer.getDataIndexCount());
+ reader.setDataBlockIndexReader(dataIndexReader);
+ // Meta index.
+ this.metaIndexReader = new
HFileBlockIndex.ByteArrayKeyBlockIndexReader(1);
+
metaIndexReader.readRootIndex(blockIter.nextBlockWithBlockType(BlockType.ROOT_INDEX),
trailer.getMetaIndexCount());
- reader.setMetaBlockIndexReader(metaIndexReader);
- loadMetaInfo(blockIter, hfileContext);
-
reader.setDataBlockEncoder(HFileDataBlockEncoderImpl.createFromFileInfo(this));
- // Load-On-Open info
- HFileBlock b;
- while ((b = blockIter.nextBlock()) != null) {
- loadOnOpenBlocks.add(b);
+ reader.setMetaBlockIndexReader(metaIndexReader);
+ loadMetaInfo(blockIter, hfileContext);
+
reader.setDataBlockEncoder(HFileDataBlockEncoderImpl.createFromFileInfo(this));
+ // Load-On-Open info
+ HFileBlock b;
+ while ((b = blockIter.nextBlock()) != null) {
+ loadOnOpenBlocks.add(b);
+ }
+ } catch (Throwable t) {
+ IOUtils.closeQuietly(context.getInputStreamWrapper());
+ throw new CorruptHFileException("Problem reading data index and meta
index from file "
+ + context.getFilePath(), t);
+ } finally {
+ context.getInputStreamWrapper().unbuffer();
Review comment:
I think it might be better to call
`context.getInputStreamWrapper().unbuffer()` before `catch` sentences.
And because `DFSInputStream.close()` uses the same logic as `unbuffer()` to
close the `blockReader` by `closeCurrentBlockReaders()`, so it need not to call
`unbuffer()` before `IOUtils.closeQuietly()`. PR has been updated.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
