Skip to content

HDDS-5483. Validate block file length during put block #2463

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mukul1987 merged 2 commits into from
Aug 2, 2021
Merged

HDDS-5483. Validate block file length during put block #2463

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ec9bbff
rebase HDDS-5483
Aug 2, 2021
d9d1778
trigger new CI check
Aug 2, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 18 additions & 1 deletion ...ner-service/src/main/java/org/apache/hadoop/ozone/container/keyvalue/KeyValueHandler.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@

package org.apache.hadoop.ozone.container.keyvalue;

import java.io.File;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
Expand Down Expand Up @@ -69,6 +70,7 @@
import org.apache.hadoop.ozone.container.common.volume.VolumeSet;
import org.apache.hadoop.ozone.container.keyvalue.helpers.BlockUtils;
import org.apache.hadoop.ozone.container.keyvalue.impl.BlockManagerImpl;
import org.apache.hadoop.ozone.container.keyvalue.impl.ChunkManagerDummyImpl;
import org.apache.hadoop.ozone.container.keyvalue.impl.ChunkManagerFactory;
import org.apache.hadoop.ozone.container.keyvalue.interfaces.BlockManager;
import org.apache.hadoop.ozone.container.keyvalue.interfaces.ChunkManager;
Expand Down Expand Up @@ -98,6 +100,7 @@
import static org.apache.hadoop.hdds.scm.protocolPB.ContainerCommandResponseBuilders.putBlockResponseSuccess;
import static org.apache.hadoop.hdds.scm.protocolPB.ContainerCommandResponseBuilders.unsupportedRequest;
import static org.apache.hadoop.hdds.scm.utils.ClientCommandsUtils.getReadChunkVersion;
import static org.apache.hadoop.ozone.container.common.impl.ChunkLayOutVersion.FILE_PER_BLOCK;

import org.apache.ratis.thirdparty.com.google.protobuf.ByteString;
import org.slf4j.Logger;
Expand Down Expand Up @@ -460,6 +463,20 @@ ContainerCommandResponseProto handlePutBlock(
ContainerProtos.BlockData data = request.getPutBlock().getBlockData();
BlockData blockData = BlockData.getFromProtoBuf(data);
Preconditions.checkNotNull(blockData);
ChunkLayOutVersion layoutVersion =
ChunkLayOutVersion.getConfiguredVersion(conf);
if (layoutVersion == FILE_PER_BLOCK &&
// ChunkManagerDummyImpl doesn't persist to disk, don't check here
!chunkManager.getClass()
.isAssignableFrom(ChunkManagerDummyImpl.class)) {
Comment on lines +468 to +471
Copy link
Contributor

@adoroszlai adoroszlai Aug 17, 2021 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think layout-specific code would be better added to each ChunkManager implementation instead of such == and isAssignableFrom checks. Take a look at finishWriteChunks() and shutdown().

File blockFile = layoutVersion
.getChunkFile(kvContainer.getContainerData(),
blockData.getBlockID(), null);
// ensure that the putBlock length <= blockFile length
Preconditions.checkArgument(blockFile.length() >= blockData.getSize(),
Copy link
Contributor

@smengcl smengcl Jul 28, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Under what conditions could this be violated?

Copy link
Contributor Author

@sadanand48 sadanand48 Aug 2, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@smengcl This is just a safety check. The ozone streaming write path will not use the current ratisAsyncApi which ensures sequential transactions i.e putBlock would always occur after writeChunk. To be on a safer side, this check is added. We may remove this after the complete streaming implementation if this check is redundant
cc : @mukul1987

"BlockData in putBlock() has more length "
+ "than the actual data written into the disk");
}

boolean incrKeyCount = false;
if (!request.getPutBlock().hasEof() || request.getPutBlock().getEof()) {
Expand All @@ -478,7 +495,7 @@ ContainerCommandResponseProto handlePutBlock(
metrics.incContainerBytesStats(Type.PutBlock, numBytes);
} catch (StorageContainerException ex) {
return ContainerUtils.logAndReturnError(LOG, ex, request);
} catch (IOException ex) {
} catch (IOException | IllegalArgumentException ex) {
return ContainerUtils.logAndReturnError(LOG,
new StorageContainerException("Put Key failed", ex, IO_EXCEPTION),
request);
Expand Down