Opened 17 years ago
Last modified 11 years ago
#232 assigned defect
Owner of symbolic links is not restored correctly.
Reported by: | enakai | Owned by: | Bruno Cornec |
---|---|---|---|
Priority: | low | Milestone: | |
Component: | mondo | Version: | 2.2.4 |
Severity: | normal | Keywords: | |
Cc: |
Description
I have restored RHEL5.1 with mondo-2.2.4-1.rhel5 and compared "rpm -Va" results taken before and after the restoration. And I found that ower user and group of symbolic links originally different from root.root was changed to root.root.
# diff node01.2008011663155/rpm-Va node01.2008011684617/rpm-Va 1218a1219,1225 > .....UG. /usr/bin/rpmdb > .....UG. /usr/bin/rpmquery > .....UG. /usr/bin/rpmsign > .....UG. /usr/bin/rpmverify > .....UG. /usr/lib/rpm/rpme > .....UG. /usr/lib/rpm/rpmu > .....UG. /usr/lib/rpm/rpmv
These are all symbolic links to another file, for example;
<Original> # ls -l /usr/bin/rpmdb lrwxrwxrwx 1 rpm rpm 15 Jan 16 17:29 /usr/bin/rpmdb -> ../lib/rpm/rpmd <After Restoration> # ls -l /usr/bin/rpmdb lrwxrwxrwx 1 root root 15 Jan 16 17:29 /usr/bin/rpmdb -> ../lib/rpm/rpmd
As owners of linked "real" files were not modified, there's probably no practical issue. But I think, if possible, it's better to fix it. (Sorry for that I have no idea how to fix it.)
RPMs I used are as below;
afio-2.4.7-1.x86_64.rpm buffer-1.19-1.x86_64.rpm mindi-1.2.4-1.rhel5.x86_64.rpm mindi-busybox-1.2.2-3.rhel5.x86_64.rpm mondo-2.2.4-1.rhel5.x86_64.rpm
Change History (7)
comment:1 by , 17 years ago
Milestone: | 2.2.5 → 2.2.6 |
---|---|
Priority: | normal → low |
comment:2 by , 17 years ago
Milestone: | 2.2.6 → 2.2.7 |
---|
comment:3 by , 16 years ago
Milestone: | 2.2.7 → 2.2.8 |
---|
comment:4 by , 16 years ago
Status: | new → assigned |
---|
comment:5 by , 16 years ago
Milestone: | 2.2.8 → 2.2.9 |
---|
comment:6 by , 16 years ago
Milestone: | 2.2.9 → 2.2.10 |
---|
Note:
See TracTickets
for help on using tickets.
afio does support restoration of owner/group id correctly. I made successful tests with that. I need to see if any option we use, does change that, but I've not found anything obvious up to now.
Will try to reproduce and debug.