Skip to content
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.

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

HBASE-27619 Bulkload fails when trying to bulkload files with invalid names after HBASE-26707 #5014

Merged
merged 2 commits into from
Feb 9, 2023
Merged
Show file tree
Hide file tree
Changes from 1 commit
Commits
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
Original file line number Diff line number Diff line change
Expand Up @@ -7280,7 +7280,7 @@ public Map<byte[], List<Path>> bulkLoadHFiles(Collection<Pair<byte[], String>> f
reqTmp ? null : fs.getRegionDir().toString());
}
Pair<Path, Path> pair = null;
if (reqTmp) {
if (reqTmp || (!StoreFileInfo.isHFile(finalPath))) {
Copy link
Contributor

Choose a reason for hiding this comment

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

nit: maybe remove the extra () ?

Suggested change
if (reqTmp || (!StoreFileInfo.isHFile(finalPath))) {
if (reqTmp || !StoreFileInfo.isHFile(finalPath)) {

do you have any stacktrace and example of the non-HFile ? also , when you said has been split in the bulkload process , what would this split generate?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

do you have any stacktrace and example of the non-HFile ?

For the case of split files, BulkLoadHFilesTool splits the file into two, naming the resulting files with " .bottom/.top" suffixes here. In this case, when using SFT, we'll put these files into the store dir, then when StoreFileReader tries to load it, it will try to resolve it as a reference, giving the following errors:

  1. At the bulkload job side:

ERROR org.apache.hadoop.hbase.regionserver.HRegion: There was a partial failure due to IO when attempting to load CF : s3a://my-s3-bucket/my-datalake/hbase/data/default/my-table/2462397a71fa06cff0c5598434a02c5f/CF/e245ce23ae9b48ebac7713568ce3a569.top

  1. At RS side:
    FileNotFoundException: No such file or directory: s3a://my-s3-bucket/my-datalake/hbase/data/default/my-table/top/CF/e245ce23ae9b48ebac7713568ce3a569

Note in the path of error #2 that the RS assumed the file was a reference and tried tor resolve it as such.

There's also the possibility clients pass files with invalid naming format. These would also be placed in the store dir as it is, then failing the validations performed by StoreFileReader during file loading. These would manifest such as below, on RS logs:

2023-02-08T10:27:26,447 ERROR [RpcServer.default.FPBQ.Fifo.handler=2,queue=0,port=49900] regionserver.HRegion(7346): There was a partial failure due to IO when attempting to load myfam : s3a://my-s3-bucket/my-datalake/hbase/data/default/my-table/3c91abacbd1a6c41d7247bd3d5a6bb00/CF/hfile_0 java.io.IOException: path=s3a://my-s3-bucket/my-datalake/hbase/data/default/my-table/3c91abacbd1a6c41d7247bd3d5a6bb00/CF/hfile_0 doesn't look like a valid StoreFile

These two are also easily reproducible by the UT added here.

pair = store.preBulkLoadHFile(finalPath, seqId);
} else {
Path livePath = new Path(finalPath);
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,55 @@
/*
* 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.hbase.tool;

import static org.apache.hadoop.hbase.regionserver.storefiletracker.StoreFileTrackerFactory.TRACKER_IMPL;

import org.apache.hadoop.hbase.HBaseClassTestRule;
import org.apache.hadoop.hbase.HConstants;
import org.apache.hadoop.hbase.codec.KeyValueCodecWithTags;
import org.apache.hadoop.hbase.coprocessor.CoprocessorHost;
import org.apache.hadoop.hbase.regionserver.storefiletracker.StoreFileTrackerFactory;
import org.apache.hadoop.hbase.testclassification.LargeTests;
import org.apache.hadoop.hbase.testclassification.MiscTests;
import org.junit.BeforeClass;
import org.junit.ClassRule;
import org.junit.experimental.categories.Category;

/**
* Test cases for LoadIncrementalHFiles when SFT is enabled.
*/
@Category({ MiscTests.class, LargeTests.class })
public class TestBulkLoadHFilesSFT extends TestBulkLoadHFiles {

@ClassRule
public static final HBaseClassTestRule CLASS_RULE =
HBaseClassTestRule.forClass(TestBulkLoadHFilesSFT.class);

@BeforeClass
public static void setUpBeforeClass() throws Exception {
util.getConfiguration().set(CoprocessorHost.REGION_COPROCESSOR_CONF_KEY, "");
util.getConfiguration().setInt(BulkLoadHFiles.MAX_FILES_PER_REGION_PER_FAMILY,
MAX_FILES_PER_REGION_PER_FAMILY);
util.getConfiguration().set(TRACKER_IMPL, StoreFileTrackerFactory.Trackers.FILE.name());
// change default behavior so that tag values are returned with normal rpcs
util.getConfiguration().set(HConstants.RPC_CODEC_CONF_KEY,
KeyValueCodecWithTags.class.getCanonicalName());
util.startMiniCluster();
setupNamespace();
}
}