Page MenuHomePhabricator

Check home/HDFS leftovers of jly
Closed, ResolvedPublic

Description

The access for Jimmy Ly (jly) has been removed. It needs to be checked if data was left in home dirs on stat*/HDFS since they were part of the "analytics-privatedata-users" group.

The Kerberos principal has already been removed.

Event Timeline

Gehel triaged this task as Low priority.Sep 17 2025, 7:40 AM

No files found in HDFS for user jly

stevemunene@an-master1003:~$ sudo kerberos-run-command hdfs hdfs dfs -ls /user/jly
stevemunene@an-master1003:~$

However, we do have some files on the stat1008 host. @OKryva-WMF kindly advise on whether we should keep/move the files on /home/jly or remove them, thanks.

stevemunene@cumin1003:~$ sudo cumin 'stat1*.eqiad.wmnet' 'ls /home/jly'
4 hosts will be targeted:
stat[1008-1011].eqiad.wmnet
OK to proceed on 4 hosts? Enter the number of affected hosts to confirm or "q" to quit: 4
===== NODE GROUP =====                                                                                                                                                                                                                                                                                          
(1) stat1008.eqiad.wmnet                                                                                                                                                                                                                                                                                        
----- OUTPUT of 'ls /home/jly' -----                                                                                                                                                                                                                                                                            
2025-03-26_account_compromises.log                                                                                                                                                                                                                                                                              
compromised_wiki_user_groups.csv                                                                                                                                                                                                                                                                                
default_queries.ipynb
hibp_compromised_accounts.csv
hibp_compromised_accounts_old.csv
hibp.ipynb
old
user_emails.csv
user_emails_high_privs_short.csv
usernames.csv
wikimedia_domains.csv
================                                                                                                                                                                                                                                                                                                
PASS |██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 100% (4/4) [00:00<00:00,  7.90hosts/s]
FAIL |                                                                                                                                                                                                                                                                          |   0% (0/4) [00:00<?, ?hosts/s]
100.0% (4/4) success ratio (>= 100.0% threshold) for command: 'ls /home/jly'.
100.0% (4/4) success ratio (>= 100.0% threshold) of nodes successfully executed all commands.

@Stevemunene : can you make those files available to @SLopes-WMF ? He can confirm when we can remove them from stat machines

In communication with @SLopes-WMF and the files have been shared for review.
Thanks @Gehel

@Stevemunene gave me a copy of the Jimmy's files. @Gehel, feel free to delete them now.

Thanks @SLopes-WMF the files have been cleared and we can close the task.

stevemunene@cumin1003:~$ sudo cumin 'stat1*.eqiad.wmnet' 'ls /home/jly'
4 hosts will be targeted:
stat[1008-1011].eqiad.wmnet
OK to proceed on 4 hosts? Enter the number of affected hosts to confirm or "q" to quit: 4
===== NO OUTPUT =====                                                                                                                                              
PASS |█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 100% (4/4) [00:00<00:00,  7.84hosts/s]
FAIL |                                                                                                                             |   0% (0/4) [00:00<?, ?hosts/s]
100.0% (4/4) success ratio (>= 100.0% threshold) for command: 'ls /home/jly'.
100.0% (4/4) success ratio (>= 100.0% threshold) of nodes successfully executed all commands.