[ 
https://issues.apache.org/jira/browse/FLINK-6842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16076515#comment-16076515
 ] 

ASF GitHub Bot commented on FLINK-6842:
---------------------------------------

Github user greghogan commented on a diff in the pull request:

    https://github.com/apache/flink/pull/4219#discussion_r125902757
  
    --- Diff: 
flink-runtime/src/main/java/org/apache/flink/runtime/fs/hdfs/HadoopFileSystem.java
 ---
    @@ -459,17 +459,14 @@ public boolean isDistributedFS() {
        @Override
        public Class<?> getHadoopWrapperClassNameForFileSystem(String scheme) {
                Configuration hadoopConf = getHadoopConfiguration();
    -           Class<? extends org.apache.hadoop.fs.FileSystem> clazz;
    -           // We can activate this block once we drop Hadoop1 support 
(only hd2 has the getFileSystemClass-method)
    -//         try {
    -//                 clazz = 
org.apache.hadoop.fs.FileSystem.getFileSystemClass(scheme, hadoopConf);
    -//         } catch (IOException e) {
    -//                 LOG.info("Flink could not load the Hadoop File system 
implementation for scheme "+scheme);
    -//                 return null;
    -//         }
    -           clazz = hadoopConf.getClass("fs." + scheme + ".impl", null, 
org.apache.hadoop.fs.FileSystem.class);
    -
    -           if (clazz != null && LOG.isDebugEnabled()) {
    +           Class<? extends org.apache.hadoop.fs.FileSystem> clazz = null;
    +           try {
    +                   clazz = 
org.apache.hadoop.fs.FileSystem.getFileSystemClass(scheme, hadoopConf);
    +           } catch (IOException e) {
    +                   LOG.info("Flink could not load the Hadoop File system 
implementation for scheme " + scheme);
    +           }
    +
    +           if(LOG.isDebugEnabled()) {
    --- End diff --
    
    Checkstyle is `if (...`. When the `IOException` is caught `clazz` will 
still be `null`.


> Uncomment or remove code in HadoopFileSystem
> --------------------------------------------
>
>                 Key: FLINK-6842
>                 URL: https://issues.apache.org/jira/browse/FLINK-6842
>             Project: Flink
>          Issue Type: Improvement
>          Components: Local Runtime
>    Affects Versions: 1.4.0
>            Reporter: Chesnay Schepler
>            Assignee: mingleizhang
>            Priority: Minor
>
> I've found the following code in 
> {{HadoopFileSystem#getHadoopWrapperClassNameForFileSystem}}
> {code}
> Configuration hadoopConf = getHadoopConfiguration();
> Class<? extends org.apache.hadoop.fs.FileSystem> clazz;
> // We can activate this block once we drop Hadoop1 support (only hd2 has the 
> getFileSystemClass-method)
> //try {
> //    clazz = org.apache.hadoop.fs.FileSystem.getFileSystemClass(scheme, 
> hadoopConf);
> //} catch (IOException e) {
> //    LOG.info("Flink could not load the Hadoop File system implementation 
> for scheme "+scheme);
> //    return null;
> //}
> clazz = hadoopConf.getClass("fs." + scheme + ".impl", null, 
> org.apache.hadoop.fs.FileSystem.class);
> if (clazz != null && LOG.isDebugEnabled()) {
>       LOG.debug("Flink supports {} with the Hadoop file system wrapper, impl 
> {}", scheme, clazz);
> }
> return clazz;
> {code}
> Since we don't support hadoop1 anymore the commented code should either be 
> activated or removed.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to