On Fri, Mar 26, 2004 at 07:48:19PM -0600, Robert wrote:
[root@mavis root]# /usr/sbin/tmpwatch 240 /tmp
and I *still* have a bunch of old files in /tmp
[root@mavis root]# find /tmp -ctime +10 | wc -l 613
....
Y'know, I bet this is gonna turn out to be something really simple that I'm overlooking -- something in the same league as the infamous logrotate problem a few years ago that caused the supply of inodes to dry up in short order.
Try "stat" on these files and dirs there are four interesting dates when thinking about files.
Creation: Access: Modify: Change:
I believe that "tmpwatch" is paying attention to the access time stamp. So if you inspect a file you are accessing it (cat, wc, virus-scanner...) and the access time will be reset.
Is that what is going on?
An tmpwatch interaction with the 'access' time and dirs makes sense because of the checking for contents thus:
mkdir /tmp/AAA
stat /tmp/AAA
sleep 60
stat /tmp/AAA
ls /tmp/AAA
stat /tmp/AAA
YES! Using the test file I created...
[root@mavis root]# touch -a -t 198506241213.14 /tmp/zzzA-test-file.txt [root@mavis root]# stat /tmp/zzzA-test-file.txt File: `/tmp/zzzA-test-file.txt' Size: 24 Blocks: 8 IO Block: 4096 regular file Device: 302h/770d Inode: 214418 Links: 1 Access: (0666/-rw-rw-rw-) Uid: ( 0/ root) Gid: ( 0/ root) Access: 1985-06-24 12:13:14.000000000 -0500 Modify: 2003-11-14 11:11:37.000000000 -0600 Change: 2004-03-27 06:20:10.000000000 -0600
[root@mavis root]# f-prot /tmp/zzzA-test-file.txt Virus scanning report - 27 March 2004 @ 6:20
F-PROT ANTIVIRUS Program version: 4.3.1 Engine version: 3.14.7
VIRUS SIGNATURE FILES SIGN.DEF created 26 March 2004 SIGN2.DEF created 26 March 2004 MACRO.DEF created 24 March 2004
Search: /tmp/zzzA-test-file.txt Action: Report only Files: Attempt to identify files Switches: <none>
Results of virus scanning:
Files: 1 MBRs: 0 Boot sectors: 0 Objects scanned: 1
Time: 0:00
No viruses or suspicious files/boot sectors were found. [root@mavis root]# stat /tmp/zzzA-test-file.txt File: `/tmp/zzzA-test-file.txt' Size: 24 Blocks: 8 IO Block: 4096 regular file Device: 302h/770d Inode: 214418 Links: 1 Access: (0666/-rw-rw-rw-) Uid: ( 0/ root) Gid: ( 0/ root) Access: 2004-03-27 06:20:40.000000000 -0600 Modify: 2003-11-14 11:11:37.000000000 -0600 Change: 2004-03-27 06:20:10.000000000 -0600
So, f-prot's access is changing the access time and the following shows that tmpwatch is working properly:
[root@mavis root]# touch -a -t 198506241213.14 /tmp/zzzA-test-file.txt [root@mavis root]# stat /tmp/zzzA-test-file.txt File: `/tmp/zzzA-test-file.txt' Size: 24 Blocks: 8 IO Block: 4096 regular file Device: 302h/770d Inode: 214418 Links: 1 Access: (0666/-rw-rw-rw-) Uid: ( 0/ root) Gid: ( 0/ root) Access: 1985-06-24 12:13:14.000000000 -0500 Modify: 2003-11-14 11:11:37.000000000 -0600 Change: 2004-03-27 06:30:30.000000000 -0600
[root@mavis root]# /usr/sbin/tmpwatch 240 /tmp [root@mavis root]# stat /tmp/zzzA-test-file.txt stat: cannot stat `/tmp/zzzA-test-file.txt': No such file or directory [root@mavis root]#
Thanks to all who responded and hammered this through my thick old head.
-- People will buy anything that's one to a customer.
06:33:01 up 2 days, 19:11, 13 users, load average: 0.00, 0.03, 0.00
One billion seconds ago it was 05:46:21 CDT Wed 07/19/72Repeat after me: "The primary purpose of any government entity is to employ the unemployable."
