[
https://issues.apache.org/jira/browse/DRILL-5365?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16548318#comment-16548318
]
ASF GitHub Bot commented on DRILL-5365:
---------------------------------------
sohami commented on a change in pull request #1296: DRILL-5365: Prevent plugin
config from changing default fs. Make DrillFileSystem Immutable.
URL: https://github.com/apache/drill/pull/1296#discussion_r203504583
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/store/dfs/DrillFileSystemCache.java
##########
@@ -0,0 +1,99 @@
+/*
+ * 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.drill.exec.store.dfs;
+
+import com.google.common.base.Preconditions;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.FileSystem;
+
+import java.io.IOException;
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * <h4>Motivation</h4>
+ * <p>
+ * This cache is intended to work around the bugs in the {@link
org.apache.hadoop.fs.FileSystem} static cache (DRILL-5365). Specifically, as of
Hadoop 2.7.x the
+ * {@link org.apache.hadoop.fs.FileSystem} cache has the following bad
behavior:
+ * </p>
+ * <ul>
+ * <li>
+ * The {@link org.apache.hadoop.conf.Configuration} object is not
considered when constructing keys for the {@link
org.apache.hadoop.fs.FileSystem} cache of
+ * {@link org.apache.hadoop.fs.FileSystem} objects.
+ * </li>
+ * <li>
+ * The {@link org.apache.hadoop.fs.FileSystem} cache does not honor the
<b>fs.default.name</b> property when constructing keys, only
<b>fs.defaultFS</b> is used to construct
+ * keys in the cache.
+ * </li>
+ * </ul>
+ *
+ * <h4>Usage</h4>
+ *
+ * <ul>
+ * <li>
+ * A prerequisite for usage is that all {@link
org.apache.hadoop.conf.Configuration} objects are normalized with
+ * {@link
org.apache.drill.exec.store.dfs.DrillFileSystem#normalize(Configuration)}.
+ * </li>
+ * <li>
+ * This cache should only be used from {@link
org.apache.drill.exec.store.dfs.DrillFileSystem}.
+ * </li>
+ * </ul>
+ *
+ * <h4>TODO</h4>
+ *
+ * <ul>
+ * <li>
+ * Drill currently keeps a {@link
org.apache.drill.exec.store.dfs.DrillFileSystem} open indefinitely. This will
be corrected
+ * in DRILL-6608. As a result this cache currently has no methods to
remove {@link org.apache.hadoop.fs.FileSystem} objects
+ * after they are created.
+ * </li>
+ * </ul>
+ */
+class DrillFileSystemCache {
+ private Map<Map<String, String>, FileSystem> cache = new HashMap<>();
+
+ /**
+ * If a {@link org.apache.hadoop.fs.FileSystem} object corresponding to the
given {@link org.apache.hadoop.conf.Configuration}
+ * exists in the cache, then it is returned. If no corresponding {@link
org.apache.hadoop.fs.FileSystem} exist, then it is created,
+ * added to the cache, and returned.
+ * @param configuration The {@link org.apache.hadoop.conf.Configuration}
corresponding to the desired {@link org.apache.hadoop.fs.FileSystem}
+ * object. It is expected that this configuration is
first normalized with {@link
org.apache.drill.exec.store.dfs.DrillFileSystem#normalize(Configuration)}.
+ * @return The {@link org.apache.hadoop.fs.FileSystem} object corresponding
to he given {@link org.apache.hadoop.conf.Configuration}.
+ * @throws IOException An error when creating the desired {@link
org.apache.hadoop.fs.FileSystem} object.
+ */
+ protected synchronized FileSystem get(final Configuration configuration)
throws IOException {
+ final Map<String, String> map = new HashMap<>(configToMap(configuration));
+
+ if (!cache.containsKey(map)) {
+ cache.put(map, FileSystem.newInstance(configuration));
+ }
+
+ return cache.get(map);
+ }
+
+ static Map<String, String> configToMap(final Configuration configuration) {
+ Preconditions.checkNotNull(configuration);
+ final Map<String, String> map = new HashMap<>();
+
+ for (Map.Entry<String, String> entry: configuration) {
+ map.put(entry.getKey().trim(), entry.getValue());
Review comment:
In this implementation we are only considering `Configuration` object in
Cache key. How about other parameters which [HadoopFileSystem
cache](https://github.com/apache/hadoop/blob/branch-2.7.1/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/FileSystem.java#L2812)
is using ? Specially I am concerned about the UGI object from which the user
information is retrieved which is later checked for access control.
Also for handling `fs.default.name` and `fs.defaultFS` once it's set in the
configuration parameter and FileSystem instance is created then
[FileSystemCache does use information from there based on
UR](https://github.com/apache/hadoop/blob/branch-2.7.1/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/FileSystem.java#L2806)I
(which is created using those parameters).
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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
> FileNotFoundException when reading a parquet file
> -------------------------------------------------
>
> Key: DRILL-5365
> URL: https://issues.apache.org/jira/browse/DRILL-5365
> Project: Apache Drill
> Issue Type: Bug
> Components: Storage - Hive
> Affects Versions: 1.10.0
> Reporter: Chun Chang
> Assignee: Timothy Farkas
> Priority: Major
> Fix For: 1.14.0
>
>
> The parquet file is generated through the following CTAS.
> To reproduce the issue: 1) two or more nodes cluster; 2) enable
> impersonation; 3) set "fs.default.name": "file:///" in hive storage plugin;
> 4) restart drillbits; 5) as a regular user, on node A, drop the table/file;
> 6) ctas from a large enough hive table as source to recreate the table/file;
> 7) query the table from node A should work; 8) query from node B as same user
> should reproduce the issue.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
