You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by "Dhirendra Khanka (JIRA)" <ji...@apache.org> on 2018/02/07 15:31:00 UTC

[jira] [Reopened] (HADOOP-15213) JniBasedUnixGroupsNetgroupMapping.java and ShellBasedUnixGroupsNetgroupMapping.java use netgroup.substring(1)

     [ https://issues.apache.org/jira/browse/HADOOP-15213?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dhirendra Khanka reopened HADOOP-15213:
---------------------------------------

> JniBasedUnixGroupsNetgroupMapping.java and ShellBasedUnixGroupsNetgroupMapping.java use netgroup.substring(1) 
> --------------------------------------------------------------------------------------------------------------
>
>                 Key: HADOOP-15213
>                 URL: https://issues.apache.org/jira/browse/HADOOP-15213
>             Project: Hadoop Common
>          Issue Type: Improvement
>          Components: security
>         Environment: SUSE Linux Enterprise Server 11 (x86_64)
> VERSION = 11
> PATCHLEVEL = 3
>            Reporter: Dhirendra Khanka
>            Priority: Minor
>
>  
> Part of the code below shown from below 2 classes
>  org.apache.hadoop.security.JniBasedUnixGroupsNetgroupMapping.java
> {code:java}
>  protected synchronized List<String> getUsersForNetgroup(String netgroup) {
>     String[] users = null;
>     try {
>       // JNI code does not expect '@' at the begining of the group name
>       users = getUsersForNetgroupJNI(netgroup.substring(1));
>     } catch (Exception e) {
>       if (LOG.isDebugEnabled()) {
>         LOG.debug("Error getting users for netgroup " + netgroup, e);
>       } else {
>         LOG.info("Error getting users for netgroup " + netgroup + 
>             ": " + e.getMessage());
>       }
>     }
>     if (users != null && users.length != 0) {
>       return Arrays.asList(users);
>     }
>     return new LinkedList<String>();
>   }{code}
> org.apache.hadoop.security.ShellBasedUnixGroupsNetgroupMapping.java
>  
> {code:java}
> protected String execShellGetUserForNetgroup(final String netgroup)
>  throws IOException {
>  String result = "";
>  try
> { // shell command does not expect '@' at the begining of the group name 
> result = Shell.execCommand( Shell.getUsersForNetgroupCommand(netgroup.substring(1))); }
> catch (ExitCodeException e)
> { // if we didn't get the group - just return empty list; LOG.warn("error getting users for netgroup " + netgroup, e); }
> return result;
>  }
> {code}
>  The comments from the code above expect the input to contain '@' , however when executing the shell directly the output has the below form which does not contain any ampersand symbol. 
> {code:java}
> :~> getent netgroup mynetgroup1
> mynetgroup1           ( , a3xsds, ) ( , beekvkl, ) ( , redcuan, ) ( , uedfmst, ){code}
>  
> I have created a test code and removed the substring function and then ran it on the cluster using hadoop jar. The code returned netgroups correctly after the modification. I have limited knowledge on netgroup. The issue was discovered when
> hadoop.security.group.mapping = *org.apache.hadoop.security.JniBasedUnixGroupsMappingWithFallback* was added to core-site.xml and it failed to apply netgroup access.
>  
>  
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: common-dev-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-dev-help@hadoop.apache.org