[
https://issues.apache.org/jira/browse/HADOOP-19232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17922145#comment-17922145
]
ASF GitHub Bot commented on HADOOP-19232:
-----------------------------------------
rakeshadr commented on code in PR #7272:
URL: https://github.com/apache/hadoop/pull/7272#discussion_r1934191424
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsOutputStream.java:
##########
@@ -237,8 +427,12 @@ public synchronized void write(final byte[] data, final
int off, final int lengt
if (hasLease() && isLeaseFreed()) {
throw new PathIOException(path, ERR_WRITE_WITHOUT_LEASE);
}
- DataBlocks.DataBlock block = createBlockIfNeeded();
- int written = block.write(data, off, length);
+ if (length == 0) {
+ return;
Review Comment:
Please add a log message conveying this condition
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsOutputStream.java:
##########
@@ -338,9 +531,14 @@ private void uploadBlockAsync(DataBlocks.DataBlock
blockToUpload,
*/
AppendRequestParameters reqParams = new AppendRequestParameters(
offset, 0, bytesLength, mode, false, leaseId,
isExpectHeaderEnabled);
- AbfsRestOperation op = getClient().append(path,
- blockUploadData.toByteArray(), reqParams, cachedSasToken.get(),
- contextEncryptionAdapter, new TracingContext(tracingContext));
+ AbfsRestOperation op;
+ try {
+ op = remoteWrite(blockToUpload, blockUploadData, reqParams,
tracingContext);
+ } catch (InvalidIngressServiceException ex) {
+ switchHandler();
Review Comment:
Whether this is deterministic switching ? I meant, in case of exception what
will be the behavior of the new handler?
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsBlobClient.java:
##########
@@ -328,7 +337,7 @@ public AbfsRestOperation listPath(final String
relativePath, final boolean recur
abfsUriQueryBuilder.addQuery(QUERY_PARAM_DELIMITER, FORWARD_SLASH);
}
abfsUriQueryBuilder.addQuery(QUERY_PARAM_MAX_RESULTS,
String.valueOf(listMaxResults));
- appendSASTokenToQuery(relativePath, SASTokenProvider.LIST_OPERATION,
abfsUriQueryBuilder);
+ appendSASTokenToQuery(relativePath,
SASTokenProvider.FIXED_SAS_STORE_OPERATION, abfsUriQueryBuilder);
Review Comment:
Hope backward compatibility being taken care?
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsOutputStream.java:
##########
@@ -631,45 +805,89 @@ private synchronized void waitForAppendsToComplete()
throws IOException {
}
}
+ /**
+ * Flushes the written bytes to the Azure Blob Storage service, ensuring all
+ * appends are completed. This method is typically called during a close
operation.
+ *
+ * @param isClose indicates whether this flush is happening as part of a
close operation.
+ * @throws IOException if an I/O error occurs during the flush operation.
+ */
private synchronized void flushWrittenBytesToService(boolean isClose) throws
IOException {
+ // Ensure all appends are completed before flushing.
waitForAppendsToComplete();
+ // Flush the written bytes to the service.
flushWrittenBytesToServiceInternal(position, false, isClose);
}
+ /**
+ * Asynchronously flushes the written bytes to the Azure Blob Storage
service.
+ * This method ensures that the write operation queue is managed and only
flushes
+ * if there are uncommitted data beyond the last flush offset.
+ *
+ * @throws IOException if an I/O error occurs during the flush operation.
+ */
private synchronized void flushWrittenBytesToServiceAsync() throws
IOException {
+ // Manage the write operation queue to ensure efficient writes
shrinkWriteOperationQueue();
+ // Only flush if there are uncommitted data beyond the last flush offset
if (this.lastTotalAppendOffset > this.lastFlushOffset) {
this.flushWrittenBytesToServiceInternal(this.lastTotalAppendOffset, true,
false/*Async flush on close not permitted*/);
}
}
+ /**
+ * Flushes the written bytes to the Azure Blob Storage service.
+ *
+ * @param offset the offset up to which data needs to be
flushed.
+ * @param retainUncommitedData whether to retain uncommitted data after
flush.
+ * @param isClose whether this flush is happening as part of a
close operation.
+ * @throws IOException if an I/O error occurs.
+ */
private synchronized void flushWrittenBytesToServiceInternal(final long
offset,
final boolean retainUncommitedData, final boolean isClose) throws
IOException {
// flush is called for appendblob only on close
if (this.isAppendBlob && !isClose) {
return;
}
+ // Tracker to monitor performance metrics
AbfsPerfTracker tracker = client.getAbfsPerfTracker();
try (AbfsPerfInfo perfInfo = new AbfsPerfInfo(tracker,
"flushWrittenBytesToServiceInternal", "flush")) {
- AbfsRestOperation op = getClient().flush(path, offset,
retainUncommitedData,
- isClose, cachedSasToken.get(), leaseId, contextEncryptionAdapter,
- new TracingContext(tracingContext));
- cachedSasToken.update(op.getSasToken());
- perfInfo.registerResult(op.getResult()).registerSuccess(true);
- } catch (AzureBlobFileSystemException ex) {
- if (ex instanceof AbfsRestOperationException) {
- if (((AbfsRestOperationException) ex).getStatusCode() ==
HttpURLConnection.HTTP_NOT_FOUND) {
+ AbfsRestOperation op;
+ try {
+ // Attempt to flush data to the remote service.
+ op = remoteFlush(offset, retainUncommitedData, isClose, leaseId,
+ tracingContext);
+ } catch (InvalidIngressServiceException ex) {
+ // If an invalid ingress service is encountered, switch handler and
retry.
+ switchHandler();
Review Comment:
The above 'switchHandler' comment is applicable here as well.
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsBlobClient.java:
##########
@@ -328,7 +337,7 @@ public AbfsRestOperation listPath(final String
relativePath, final boolean recur
abfsUriQueryBuilder.addQuery(QUERY_PARAM_DELIMITER, FORWARD_SLASH);
}
abfsUriQueryBuilder.addQuery(QUERY_PARAM_MAX_RESULTS,
String.valueOf(listMaxResults));
- appendSASTokenToQuery(relativePath, SASTokenProvider.LIST_OPERATION,
abfsUriQueryBuilder);
+ appendSASTokenToQuery(relativePath,
SASTokenProvider.FIXED_SAS_STORE_OPERATION, abfsUriQueryBuilder);
Review Comment:
Whats the plan for user delegation SAS, any jira created for this? If not
please create a follow-up jira.
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsOutputStream.java:
##########
@@ -338,9 +531,14 @@ private void uploadBlockAsync(DataBlocks.DataBlock
blockToUpload,
*/
AppendRequestParameters reqParams = new AppendRequestParameters(
offset, 0, bytesLength, mode, false, leaseId,
isExpectHeaderEnabled);
- AbfsRestOperation op = getClient().append(path,
- blockUploadData.toByteArray(), reqParams, cachedSasToken.get(),
- contextEncryptionAdapter, new TracingContext(tracingContext));
+ AbfsRestOperation op;
+ try {
+ op = remoteWrite(blockToUpload, blockUploadData, reqParams,
tracingContext);
+ } catch (InvalidIngressServiceException ex) {
+ switchHandler();
Review Comment:
Also, please add debug log message in conditional execution cases.
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AzureBlobBlockManager.java:
##########
@@ -0,0 +1,194 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.fs.azurebfs.services;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.LinkedList;
+import java.util.List;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import
org.apache.hadoop.fs.azurebfs.contracts.exceptions.AzureBlobFileSystemException;
+import org.apache.hadoop.fs.azurebfs.utils.TracingContext;
+import org.apache.hadoop.fs.store.DataBlocks;
+
+/**
+ * Manages Azure Blob blocks for append operations.
+ */
+public class AzureBlobBlockManager extends AzureBlockManager {
+
+ private static final Logger LOG = LoggerFactory.getLogger(
+ AbfsOutputStream.class);
+
+
+ /** The list of already committed blocks is stored in this list. */
+ private List<String> committedBlockEntries = new ArrayList<>();
+
+ /** The list to store blockId, position, and status. */
+ private final LinkedList<BlockEntry> blockEntryList = new LinkedList<>();
+
+
+ /**
+ * Constructs an AzureBlobBlockManager.
+ *
+ * @param abfsOutputStream the output stream
+ * @param blockFactory the block factory
+ * @param bufferSize the buffer size
+ * @throws AzureBlobFileSystemException if an error occurs
+ */
+ public AzureBlobBlockManager(AbfsOutputStream abfsOutputStream,
+ DataBlocks.BlockFactory blockFactory,
+ int bufferSize)
+ throws AzureBlobFileSystemException {
+ super(abfsOutputStream, blockFactory, bufferSize);
+ if (abfsOutputStream.getPosition() > 0 &&
!abfsOutputStream.isAppendBlob()) {
+ this.committedBlockEntries =
getBlockList(abfsOutputStream.getTracingContext());
+ }
+ LOG.debug("Created a new Blob Block Manager for AbfsOutputStream instance
{} for path {}",
+ abfsOutputStream.getStreamID(), abfsOutputStream.getPath());
+ }
+
+ /**
+ * Creates a new block.
+ *
+ * @param position the position
+ * @return the created block
+ * @throws IOException if an I/O error occurs
+ */
+ @Override
+ protected synchronized AbfsBlock createBlockInternal(long position)
+ throws IOException {
+ if (getActiveBlock() == null) {
+ setBlockCount(getBlockCount() + 1);
+ AbfsBlock activeBlock = new AbfsBlobBlock(getAbfsOutputStream(),
position);
+ activeBlock.setBlockEntry(addNewEntry(activeBlock.getBlockId(),
activeBlock.getOffset()));
+ setActiveBlock(activeBlock);
+ }
+ return getActiveBlock();
+ }
+
+ /**
+ * Returns block id's which are committed for the blob.
+ *
+ * @param tracingContext Tracing context object.
+ * @return list of committed block id's.
+ * @throws AzureBlobFileSystemException if an error occurs
+ */
+ private List<String> getBlockList(TracingContext tracingContext)
+ throws AzureBlobFileSystemException {
+ List<String> committedBlockIdList = new ArrayList<>();
+ AbfsBlobClient blobClient =
getAbfsOutputStream().getClientHandler().getBlobClient();
+ final AbfsRestOperation op = blobClient
+ .getBlockList(getAbfsOutputStream().getPath(), tracingContext);
+ if (op != null && op.getResult() != null) {
+ committedBlockIdList = op.getResult().getBlockIdList();
+ }
+ return committedBlockIdList;
+ }
+
+ /**
+ * Adds a new block entry to the block entry list.
+ * The block entry is added only if the position of the new block
+ * is greater than the position of the last block in the list.
+ *
+ * @param blockId The ID of the new block to be added.
+ * @param position The position of the new block in the stream.
+ * @return The newly added {@link BlockEntry}.
+ * @throws IOException If the position of the new block is not greater than
the last block in the list.
+ */
+ private synchronized BlockEntry addNewEntry(String blockId, long position)
throws IOException {
+ if (!blockEntryList.isEmpty()) {
+ BlockEntry lastEntry = blockEntryList.getLast();
+ if (position <= lastEntry.getPosition()) {
+ throw new IOException("New block position " + position + " must be
greater than the last block position "
+ + lastEntry.getPosition() + " for path " +
getAbfsOutputStream().getPath());
+ }
+ }
+ BlockEntry blockEntry = new BlockEntry(blockId, position,
AbfsBlockStatus.NEW);
+ blockEntryList.addLast(blockEntry);
+ LOG.debug("Added block {} at position {} with status NEW.", blockId,
position);
+ return blockEntry;
+ }
+
+ /**
+ * Updates the status of an existing block entry to SUCCESS.
+ * This method is used to mark a block as successfully processed.
+ *
+ * @param block The {@link AbfsBlock} whose status needs to be updated to
SUCCESS.
+ */
+ protected synchronized void updateEntry(AbfsBlock block) {
+ BlockEntry blockEntry = block.getBlockEntry();
+ blockEntry.setStatus(AbfsBlockStatus.SUCCESS);
+ LOG.debug("Added block {} at position {} with status SUCCESS.",
block.getBlockId(), blockEntry.getPosition());
+ }
+
+ /**
+ * Prepares the list of blocks to commit.
+ *
+ * @return whether we have some data to commit or not.
+ * @throws IOException if an I/O error occurs
+ */
+ protected boolean hasListToCommit() throws IOException {
+ // Adds all the committed blocks if available to the list of blocks to be
added in putBlockList.
+ if (blockEntryList.isEmpty()) {
+ return false; // No entries to commit
+ }
+ while (!blockEntryList.isEmpty()) {
+ BlockEntry current = blockEntryList.poll();
+ if (current.getStatus() != AbfsBlockStatus.SUCCESS) {
+ LOG.debug(
+ "Block {} with position {} has status {}, flush cannot proceed.",
+ current.getBlockId(), current.getPosition(), current.getStatus());
+ throw new IOException("Flush failed. Block " + current.getBlockId()
+ + " with position " + current.getPosition() + " has status "
+ + current.getStatus() + "for path " +
getAbfsOutputStream().getPath());
+ }
+ if (!blockEntryList.isEmpty()) {
+ BlockEntry next = blockEntryList.getFirst();
+ if (current.getPosition() >= next.getPosition()) {
+ String errorMessage =
+ "Position check failed. Current block position is greater than
or equal to the next block's position.\n"
+ + "Current Block Entry:\n"
+ + "Block ID: " + current.getBlockId()
+ + ", Position: " + current.getPosition()
+ + ", Status: " + current.getStatus()
+ + ", Path: " + getAbfsOutputStream().getPath()
+ + ", StreamID: " + getAbfsOutputStream().getStreamID()
+ + ", Next block position: " + next.getPosition()
+ + "\n";
+ throw new IOException(errorMessage);
+ }
+ }
+ committedBlockEntries.add(current.getBlockId());
Review Comment:
Please create a follow-up PR. Probably you can group all the open comments
and handle it in that PR.
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AzureBlobBlockManager.java:
##########
@@ -0,0 +1,194 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.fs.azurebfs.services;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.LinkedList;
+import java.util.List;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import
org.apache.hadoop.fs.azurebfs.contracts.exceptions.AzureBlobFileSystemException;
+import org.apache.hadoop.fs.azurebfs.utils.TracingContext;
+import org.apache.hadoop.fs.store.DataBlocks;
+
+/**
+ * Manages Azure Blob blocks for append operations.
+ */
+public class AzureBlobBlockManager extends AzureBlockManager {
+
+ private static final Logger LOG = LoggerFactory.getLogger(
+ AbfsOutputStream.class);
+
+
+ /** The list of already committed blocks is stored in this list. */
+ private List<String> committedBlockEntries = new ArrayList<>();
+
+ /** The list to store blockId, position, and status. */
+ private final LinkedList<BlockEntry> blockEntryList = new LinkedList<>();
+
+
+ /**
+ * Constructs an AzureBlobBlockManager.
+ *
+ * @param abfsOutputStream the output stream
+ * @param blockFactory the block factory
+ * @param bufferSize the buffer size
+ * @throws AzureBlobFileSystemException if an error occurs
+ */
+ public AzureBlobBlockManager(AbfsOutputStream abfsOutputStream,
+ DataBlocks.BlockFactory blockFactory,
+ int bufferSize)
+ throws AzureBlobFileSystemException {
+ super(abfsOutputStream, blockFactory, bufferSize);
+ if (abfsOutputStream.getPosition() > 0 &&
!abfsOutputStream.isAppendBlob()) {
+ this.committedBlockEntries =
getBlockList(abfsOutputStream.getTracingContext());
+ }
+ LOG.debug("Created a new Blob Block Manager for AbfsOutputStream instance
{} for path {}",
+ abfsOutputStream.getStreamID(), abfsOutputStream.getPath());
+ }
+
+ /**
+ * Creates a new block.
+ *
+ * @param position the position
+ * @return the created block
+ * @throws IOException if an I/O error occurs
+ */
+ @Override
+ protected synchronized AbfsBlock createBlockInternal(long position)
+ throws IOException {
+ if (getActiveBlock() == null) {
+ setBlockCount(getBlockCount() + 1);
+ AbfsBlock activeBlock = new AbfsBlobBlock(getAbfsOutputStream(),
position);
+ activeBlock.setBlockEntry(addNewEntry(activeBlock.getBlockId(),
activeBlock.getOffset()));
+ setActiveBlock(activeBlock);
+ }
+ return getActiveBlock();
+ }
+
+ /**
+ * Returns block id's which are committed for the blob.
+ *
+ * @param tracingContext Tracing context object.
+ * @return list of committed block id's.
+ * @throws AzureBlobFileSystemException if an error occurs
+ */
+ private List<String> getBlockList(TracingContext tracingContext)
+ throws AzureBlobFileSystemException {
+ List<String> committedBlockIdList = new ArrayList<>();
+ AbfsBlobClient blobClient =
getAbfsOutputStream().getClientHandler().getBlobClient();
+ final AbfsRestOperation op = blobClient
+ .getBlockList(getAbfsOutputStream().getPath(), tracingContext);
+ if (op != null && op.getResult() != null) {
+ committedBlockIdList = op.getResult().getBlockIdList();
+ }
+ return committedBlockIdList;
+ }
+
+ /**
+ * Adds a new block entry to the block entry list.
+ * The block entry is added only if the position of the new block
+ * is greater than the position of the last block in the list.
+ *
+ * @param blockId The ID of the new block to be added.
+ * @param position The position of the new block in the stream.
+ * @return The newly added {@link BlockEntry}.
+ * @throws IOException If the position of the new block is not greater than
the last block in the list.
+ */
+ private synchronized BlockEntry addNewEntry(String blockId, long position)
throws IOException {
+ if (!blockEntryList.isEmpty()) {
+ BlockEntry lastEntry = blockEntryList.getLast();
+ if (position <= lastEntry.getPosition()) {
+ throw new IOException("New block position " + position + " must be
greater than the last block position "
+ + lastEntry.getPosition() + " for path " +
getAbfsOutputStream().getPath());
+ }
+ }
+ BlockEntry blockEntry = new BlockEntry(blockId, position,
AbfsBlockStatus.NEW);
+ blockEntryList.addLast(blockEntry);
+ LOG.debug("Added block {} at position {} with status NEW.", blockId,
position);
+ return blockEntry;
+ }
+
+ /**
+ * Updates the status of an existing block entry to SUCCESS.
+ * This method is used to mark a block as successfully processed.
+ *
+ * @param block The {@link AbfsBlock} whose status needs to be updated to
SUCCESS.
+ */
+ protected synchronized void updateEntry(AbfsBlock block) {
+ BlockEntry blockEntry = block.getBlockEntry();
+ blockEntry.setStatus(AbfsBlockStatus.SUCCESS);
+ LOG.debug("Added block {} at position {} with status SUCCESS.",
block.getBlockId(), blockEntry.getPosition());
+ }
+
+ /**
+ * Prepares the list of blocks to commit.
+ *
+ * @return whether we have some data to commit or not.
+ * @throws IOException if an I/O error occurs
+ */
+ protected boolean hasListToCommit() throws IOException {
Review Comment:
since its protected, it can be overridden in future and can break this
assumption unknowingly. Since "synchronized" reentrant by nature, its better to
make this method also "synchronized"
> ABFS: [FnsOverBlob] Implementing Ingress Support with various Fallback
> Handling
> -------------------------------------------------------------------------------
>
> Key: HADOOP-19232
> URL: https://issues.apache.org/jira/browse/HADOOP-19232
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/azure
> Affects Versions: 3.4.0
> Reporter: Anuj Modi
> Assignee: Anmol Asrani
> Priority: Major
> Labels: pull-request-available
>
> Scope of this task is to refactor the AbfsOutputStream class to handle the
> ingress for DFS and Blob endpoint effectively.
> More details will be added soon.
> Perquisites for this Patch:
> 1. [HADOOP-19187] ABFS: [FnsOverBlob]Making AbfsClient Abstract for
> supporting both DFS and Blob Endpoint - ASF JIRA (apache.org)
> 2. [HADOOP-19226] ABFS: [FnsOverBlob]Implementing Azure Rest APIs on Blob
> Endpoint for AbfsBlobClient - ASF JIRA (apache.org)
> 3. [HADOOP-19207] ABFS: [FnsOverBlob]Response Handling of Blob Endpoint APIs
> and Metadata APIs - ASF JIRA (apache.org)
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
