Dear members,
I got the strange output on mtime information. Does anyone have such
experiences?
Could you try to do this test, please?
My Lustre environment is that client: 1.6.5.1-2.6.18_53.1.6.el5 and
OSS: 1.6.6.-2.6.18_92.1.10.el5.
Only such case of echo "">> mv_test.txt does NOT update mtime.
touch
command and new creation with echo command is OK. And, in case of
adding the time using sleep command, it also is OK.
$ cat check.sh
1 #!/bin/sh
2
3 echo "## START ORIGINAL TIME STAMP"
4 cp -p mv_test.txt.org mv_test.txt
5 #cp mv_test.txt.org mv_test.txt
6 ls -i --full-time mv_test.txt*| awk ''{print
$10"\t"$1"\t"$7" "$8}''
7 #stat mv_test.txt*
8
9 echo "## REVISE mv_test.txt"
10 #sleep 1
11 #touch mv_test.txt
12 echo "">> mv_test.txt
13 #echo ""> mv_test.txt
14
15 echo "## CHECK TIME STAMP"
16 ls -i --full-time mv_test.txt*| awk ''{print
$10"\t"$1"\t"$7" "$8}''
17
18 #rm mv_test.txt.org
19 COUNT=0
20
21 for i in `seq 1 2`
22 do
23
24 COUNT=`expr $COUNT + 1`
25
26 # touch mv_test.txt
27 mv mv_test.txt mv_test.txt.mv
28
29 echo "No.$COUNT"
30 ls -i --full-time mv_test.txt*| awk ''{print
$10"\t"$1"\t"$7" "$8}''
31
32 mv mv_test.txt.mv mv_test.txt
33
34 COUNT=`expr $COUNT + 1`
35
36 echo "No.$COUNT"
37 ls -i --full-time mv_test.txt*| awk ''{print
$10"\t"$1"\t"$7" "$8}''
38 # stat mv_test.txt*
39
40 done
Regards,
Satoshi Isono
On 2010-02-25, at 20:17, Satoshi Isono wrote:> I got the strange output on mtime information. Does anyone have such > experiences? Could you try to do this test, please? > My Lustre environment is that client: 1.6.5.1-2.6.18_53.1.6.el5 and > OSS: 1.6.6.-2.6.18_92.1.10.el5.I suggest that you upgrade to a newer version of Lustre. There were a number of mtime fixes, along with hundreds of other bug fixes since 1.6.5. Cheers, Andreas -- Andreas Dilger Sr. Staff Engineer, Lustre Group Sun Microsystems of Canada, Inc.
Dear Andreas, Thank you for your opinion. You means I should do version up on client. Does it mean from 1.6.5 to 1.6.6? By which version is it fixed? And could you do test using my script and report me with output? Regards, Satoshi Isono At 10/02/26???13:25, Andreas Dilger wrote:>On 2010-02-25, at 20:17, Satoshi Isono wrote: >>I got the strange output on mtime information. Does anyone have such >>experiences? Could you try to do this test, please? >>My Lustre environment is that client: 1.6.5.1-2.6.18_53.1.6.el5 and >>OSS: 1.6.6.-2.6.18_92.1.10.el5. > > >I suggest that you upgrade to a newer version of Lustre. There were a >number of mtime fixes, along with hundreds of other bug fixes since >1.6.5. > >Cheers, Andreas >-- >Andreas Dilger >Sr. Staff Engineer, Lustre Group >Sun Microsystems of Canada, Inc.
On 2010-02-25, at 22:52, Satoshi Isono wrote:> Thank you for your opinion. > You means I should do version up on client. Does it mean from 1.6.5 > to 1.6.6? By which version is it fixed?I would suggest 1.8.2, which is the most recent release.> And could you do test using my script and report me with output?sh check.sh ## START ORIGINAL TIME STAMP mv_test.txt 277980 2010-02-25 23:34:11.000000000 mv_test.txt.org 279316 2010-02-25 23:34:11.000000000 ## REVISE mv_test.txt ## CHECK TIME STAMP mv_test.txt 277980 2010-02-25 23:34:48.000000000 mv_test.txt.org 279316 2010-02-25 23:34:11.000000000 You are free to look through the lustre/ChangeLog to see which bug(s) contained fixes for this problem, but the above shows it is at least fixed in 1.8.2.> At 10/02/26???13:25, Andreas Dilger wrote: >> On 2010-02-25, at 20:17, Satoshi Isono wrote: >>> I got the strange output on mtime information. Does anyone have such >>> experiences? Could you try to do this test, please? >>> My Lustre environment is that client: 1.6.5.1-2.6.18_53.1.6.el5 and >>> OSS: 1.6.6.-2.6.18_92.1.10.el5. >> >> >> I suggest that you upgrade to a newer version of Lustre. There >> were a >> number of mtime fixes, along with hundreds of other bug fixes since >> 1.6.5. >> >> Cheers, Andreas >> -- >> Andreas Dilger >> Sr. Staff Engineer, Lustre Group >> Sun Microsystems of Canada, Inc. > > _______________________________________________ > Lustre-discuss mailing list > Lustre-discuss at lists.lustre.org > http://lists.lustre.org/mailman/listinfo/lustre-discussCheers, Andreas -- Andreas Dilger Sr. Staff Engineer, Lustre Group Sun Microsystems of Canada, Inc.
Thank you, Andreas If possible, can you send all output after run this script? I know that latest version is 1.8.2. Do you know 1.6.6 was able to fix this mtime problem? I would like to know LBUG number about this issue. Many thanks, Satoshi Isono At 10/02/26???15:37, Andreas Dilger wrote:>On 2010-02-25, at 22:52, Satoshi Isono wrote: >>Thank you for your opinion. >>You means I should do version up on client. Does it mean from 1.6.5 >>to 1.6.6? By which version is it fixed? > >I would suggest 1.8.2, which is the most recent release. > >>And could you do test using my script and report me with output? > >sh check.sh >## START ORIGINAL TIME STAMP >mv_test.txt 277980 2010-02-25 23:34:11.000000000 >mv_test.txt.org 279316 2010-02-25 23:34:11.000000000 >## REVISE mv_test.txt >## CHECK TIME STAMP >mv_test.txt 277980 2010-02-25 23:34:48.000000000 >mv_test.txt.org 279316 2010-02-25 23:34:11.000000000 > >You are free to look through the lustre/ChangeLog to see which bug(s) >contained fixes for this problem, but the above shows it is at least >fixed in 1.8.2. > > >>At 10/02/26???13:25, Andreas Dilger wrote: >>>On 2010-02-25, at 20:17, Satoshi Isono wrote: >>>>I got the strange output on mtime information. Does anyone have such >>>>experiences? Could you try to do this test, please? >>>>My Lustre environment is that client: 1.6.5.1-2.6.18_53.1.6.el5 and >>>>OSS: 1.6.6.-2.6.18_92.1.10.el5. >>> >>> >>>I suggest that you upgrade to a newer version of Lustre. There >>>were a >>>number of mtime fixes, along with hundreds of other bug fixes since >>>1.6.5. >>> >>>Cheers, Andreas >>>-- >>>Andreas Dilger >>>Sr. Staff Engineer, Lustre Group >>>Sun Microsystems of Canada, Inc. >> >>_______________________________________________ >>Lustre-discuss mailing list >>Lustre-discuss at lists.lustre.org >>http://lists.lustre.org/mailman/listinfo/lustre-discuss > > >Cheers, Andreas >-- >Andreas Dilger >Sr. Staff Engineer, Lustre Group >Sun Microsystems of Canada, Inc.