hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Wei Yan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-13045) RBF: Improve error message returned from subcluster
Date Fri, 06 Apr 2018 18:45:00 GMT

    [ https://issues.apache.org/jira/browse/HDFS-13045?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16428768#comment-16428768

Wei Yan commented on HDFS-13045:

I cannot produce the error pattern I mentioned above through some fs operations. Not sure
whether it is still valid. The following code is just a quick test against this error pattern.
String msg = "Parent directory doesn't exist: /a/a/b";
String src = "/ns1/a";
String dst = "/a";
String newMsg = msg.replaceAll(dst, src);
int minLen = Math.min(dst.length(), src.length());
for (int i = 0; newMsg.equals(msg) && i < minLen; i++) {
  // Check if we can replace sub folders
  String dst1 = dst.substring(0, dst.length() - 1 - i);
  String src1 = src.substring(0, src.length() - 1 - i);
  newMsg = msg.replaceAll(dst1, src1);
I think currently the patch cannot handle it, as it replaceAll at the first place.

One more nit thing is, we may also need to set the stackTrace back when generating the new

> RBF: Improve error message returned from subcluster
> ---------------------------------------------------
>                 Key: HDFS-13045
>                 URL: https://issues.apache.org/jira/browse/HDFS-13045
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>            Reporter: Wei Yan
>            Assignee: Wei Yan
>            Priority: Minor
>         Attachments: HDFS-13045.000.patch, HDFS-13045.001.patch, HDFS-13045.002.patch
> Currently, Router directly returns exception response from subcluster to client, which
may not have the correct error message, especially when the error message containing a path.
> One example, we have a mount path "/a/b" mapped to subclusterA's "/c/d". If user1 does
a chown operation on "/a/b", and he doesn't have corresponding privilege, currently the
error msg looks like "Permission denied. user=user1 is not the owner of inode=/c/d", which
may confuse user. Would be better to reverse the path back to original mount path.

This message was sent by Atlassian JIRA

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

View raw message