HDFS-17849 : Fix for Namenode crashed while cleaning up Expired Delegation tokens of older realm #8054
+5
−1
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…delegation tokens from cache and proceeding to remove
arunreddyav
changed the title
Hexiaoqiao
reviewed
Nov 4, 2025
| expiredTokens.add(entry.getKey()); | ||
| removeTokenForOwnerStats(entry.getKey()); | ||
| try { | ||
| removeTokenForOwnerStats(entry.getKey()); |
There was a problem hiding this comment.
Thanks @arunreddyav for your report and contribution, I am a little confused the token could be leak when thrown exception here. I think the smooth way is config the hadoop.security.auth_to_local when change the realm. What do you think about? Thanks again.
There was a problem hiding this comment.
- The token will not be leaked as I'm catching the exception and cleaned up in the
logExpireTokens(expiredTokens);. - Including the older rules under hadoop.security.auth_to_local could be a possible approach; however, the customer prefers not to include the older rules for security reasons (for ex :- when moved to more secure zone old keytabs should not be allowed)
There was a problem hiding this comment.
- Got it. Make sense to me. However
tokenOwnerStatscould not be clean, this is one nit issue. - 'when moved to more secure zone old keytabs should not be allowed' - I think this should be resolved at KDC side not Hadoop.
There was a problem hiding this comment.
Thanks for the review @Hexiaoqiao . Our customers create cluster does a sample job checks with local kerberoes. Once initial setup is done they will configure LDAP/Active Directory through Ambari . Once AD realm is configured they can not keep old realm based auth rules as its against the security
|
💔 -1 overall
This message was automatically generated. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description of PR
JIRA: HDFS-17849. Fix for NN crash issue during token cleanup after updating the kerb auth rules to pickup new realm configuration from existing one.
How was this patch tested?
I have tested this change on Hadoop 3.4.1 by replacing the hadoop-common JARs.
The log message appears correctly in the NameNode logs as shared below, and the NameNode starts up successfully.
For code changes:
LICENSE,LICENSE-binary,NOTICE-binaryfiles?