WARNING: Tento dopis byl pozmenen programem sendmail/MIMEDefang. This e-mail has been altered by sendmail/MIMEDefang. Priloha pojmenovana 'test.sh' byla odstranena, protoze mohla obsahovat virus. Pokud tuto prilohu skutecne potrebujete, kontaktujte odesilatele a domluvte s nim jiny zpusob zaslani (archiv RAR, apod). An attachment named 'test.sh' was removed from this email as it constituted a security hazard. If you require this attachment, please contact the sender and arrange an alternate means of receiving it (RAR archive, etc).
Hi,
I have the foloving setup:
- NFS clients machines running vanilla 2.6.18 (config included) writing on
NFS mounted share on several NFS server (inc. Lin 2.6.16.28, FreeBSD 6.1
etc).
- when two or more clients try to write/append to the same file on NFS
share this file contains all the data writen + some blocks are padded
with zeros.
- I used this test script:
#!/bin/sh
while true; do
cat /tmp/data >>/pub/tmp/zeros.log
usleep 1000;
done
- /pub is the NFS share from Lin 2.6.16.28 server
- resulting zeros.log is included
Expect it may be related to this:
http://www.ussg.iu.edu/hypermail/linux/kernel/0604.3/1469.html
([PATCH INTRO] Re: [RFC] copy_from_user races with readpage)
The same clients running 2.6.16.28 are fine.
Is this known problem (have search vanila and nfs git repos but have not
found anything reasonable) or some new bug?
If you would like more infos please let me know.
(Please CC me as I'm not in the lists.)
Regards
--
Tomas Kasparek, PhD student E-mail: [email protected]
CVT FIT VUT Brno, BI/140a Web: http://www.fit.vutbr.cz/~kasparek
Bozetechova 2, 612 66 Fax: +420 54114-1270
Brno, Czech Republic Phone: +420 54114-1220
ICQ: 293092805 jabber: [email protected]
GPG: 2F1E 1AAF FD3B CFA3 1537 63BD DCBE 18FF A035 53BC
dnetc v2.9012-497-CFR-05111007 for Linux (Linux 2.6.18). Using email address (distributed.net ID) '[email protected]' [Sep 28 06:17:51 UTC] Automatic processor type detection found an AMD K8-15 (Athlon 64) processor. [Sep 28 06:17:51 UTC] OGR-P2: using core #3 (GARSP 6.0-asm-rt1-mmx). [Sep 28 06:17:51 UTC] OGR-P2: Loaded 25/8-21-2-35-3-13 (25.00 Gnodes done) [Sep 28 06:17:52 UTC] OGR-P2: 62 packets remain in buff-in.ogf [Sep 28 06:17:52 UTC] OGR-P2: 0 packets are in buff-out.ogf [Sep 28 06:17:52 UTC] 1 cruncher has been started. [Sep 28 06:17:53 UTC] Shutdown - reached time limit. [Sep 28 06:17:53 UTC] OGR-P2: Saved 25/8-21-9-11-31-2 (19.86 Gnodes done) 0.00:08:50.35 - [24,863,350 nodes/s] [Sep 28 06:17:53 UTC] OGR-P2: Summary: 1 packet (3.30 stats units) 0.00:02:12.88 - [23.81 Mnodes/s] [Sep 28 06:17:53 UTC] OGR-P2: 63 packets are in buff-in.ogf [Sep 28 06:17:53 UTC] OGR-P2: 0 packets are in buff-out.ogf [Sep 28 06:17:53 UTC] *Break* Shutting down... [Sep 28 06:17:53 UTC] Shutdown complete. [Sep 28 06:17:54 UTC] Shutdown - reached time limit. [Sep 28 06:17:54 UTC] OGR-P2: Saved 25/8-21-9-5-31-11 (18.88 Gnodes done) 0.00:10:04.53 - [24,760,670 nodes/s] [Sep 28 06:17:54 UTC] OGR-P2: 64 packets are in buff-in.ogf [Sep 28 06:17:54 UTC] OGR-P2: 0 packets are in buff-out.ogf [Sep 28 06:17:54 UTC] *Break* Shutting down... [Sep 28 06:17:54 UTC] Shutdown complete.
Attachment:
zeros.log-bad
Description: Binary data
- Prev by Date: Re: [PATCH] Reset file->f_op in snd_card_file_remove()
- Next by Date: Re: [PATCH] Put the BUG __FILE__ and __LINE__ info out of line
- Previous by thread: Re: [PATCH] Reset file->f_op in snd_card_file_remove()
- Next by thread: specifying the order of calling kernel functions (or modules)
- Index(es):
 |