Re: Poor NFS performance after recent update
Ok just to finish this thread off: After taking the power away from my switch for 30 seconds and powering it up again - everything automagically works back to normals. Merry Christmas & a Happy new Year. -- Kimi ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
On Sat, Dec 16, 2006 at 02:21:53AM +, Kimi Ostro wrote: > Okay, this is getting stranger. transferring data between 8 machines > on my network which are all running FreeBSD as having this problem, > yet I cans download iso file off the internet at over 100KB/s. Try transfers in different directions. Example: [EMAIL PROTECTED] scp -p [EMAIL PROTECTED]:/dev/urandom /tmp/urandom [EMAIL PROTECTED] scp -p /dev/urandom [EMAIL PROTECTED]:/tmp/urandom Also try the same from machineBB: [EMAIL PROTECTED] scp -p [EMAIL PROTECTED]:/dev/urandom /tmp/urandom [EMAIL PROTECTED] scp -p /dev/urandom [EMAIL PROTECTED]:/tmp/urandom Do you see slow LAN transfer speeds when only receiving data vs. sending data? If so, chances are it's a duplex problem, and you should either adjust some settings in ifconfig_fxp0 in rc.conf to try and work around it (try forcing media 100baseTX mediaopt full-duplex rather than using autonegotiate), or try a different switch. -- | Jeremy Chadwick jdc at parodius.com | | Parodius Networkinghttp://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
Okay, this is getting stranger. transferring data between 8 machines on my network which are all running FreeBSD as having this problem, yet I cans download iso file off the internet at over 100KB/s. -- Kimi ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
On 16/12/06, Jeremy Chadwick <[EMAIL PROTECTED]> wrote: Your problem might be duplex-related. Can you provide some netstat -in output (after you've scp'd stuff, etc.), as well as ifconfig -a output? nxclient-1.4.0-91.i386.tar.gz 100% 3423KB 23.1KB/s 02:28 NameMtu Network Address Ipkts IerrsOpkts Oerrs Coll fxp0 1500 00:90:27:a4:0f:2c 3940 0 3855 0 0 fxp0 1500 192.168.0./24 192.168.0.2203962 - 3875 - - lo0 163840 00 0 0 lo0 16384 (28)00:00:00:00:00:00:fe:80:00:02:00:00:00:00:00:00:00:00:00:00:00:01 0 00 0 0 lo0 16384 (28)00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:01 0 00 0 0 lo0 16384 127 127.0.0.10 -0 - - fxp0: flags=8843 mtu 1500 options=8 inet 192.168.0.220 netmask 0xff00 broadcast 192.168.0.255 ether 00:90:27:a4:0f:2c media: Ethernet autoselect (100baseTX ) status: active lo0: flags=8049 mtu 16384 inet6 fe80::1%lo0 prefixlen 64 scopeid 0x2 inet6 ::1 prefixlen 128 inet 127.0.0.1 netmask 0xff00 looks like a normal day to me besides the 23.1KB/s 02:28 :( -- Kimi ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
On Sat, Dec 16, 2006 at 12:30:11AM +, Kimi Ostro wrote: > Looks like I was wrong, as SCP is also just as slow as NFS > > I'm lost. > > I'm going to install 6.1-RELEASE to see if that "fixes" my problem. Your problem might be duplex-related. Can you provide some netstat -in output (after you've scp'd stuff, etc.), as well as ifconfig -a output? -- | Jeremy Chadwick jdc at parodius.com | | Parodius Networkinghttp://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
Looks like I was wrong, as SCP is also just as slow as NFS I'm lost. I'm going to install 6.1-RELEASE to see if that "fixes" my problem. thanks! all!! -- Kimi ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
Kimi Ostro <[EMAIL PROTECTED]> wrote: > fserver:/data /media/datanfs > rw,-b,-i,-s,-L,noauto A small side note (probably not related to your actual problem): It is usually a bad idea to use the -s ("soft") option, because many programs are not prepared to handle unexpected I/O errors, leading to file corruption. The following is an excerpt from the Solaris docs, which applies to FreeBSD as well: | File systems that are mounted read-write or that con- | tain executable files should always be mounted with | the hard option. Applications using soft mounted file | systems may incur unexpected I/O errors, file corrup- | tion, and unexpected program core dumps. The soft | option is not recommended. I think a similar paragraph should be added to FreeBSD's mount_nfs(8) manpage and to the Handbook. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing Dienstleistungen mit Schwerpunkt FreeBSD: http://www.secnetix.de/bsd Any opinions expressed in this message may be personal to the author and may not necessarily reflect the opinions of secnetix in any way. "I have stopped reading Stephen King novels. Now I just read C code instead." -- Richard A. O'Keefe ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
On 15/12/06, Jay Chandler <[EMAIL PROTECTED]> wrote: I do run statd and lockd, but let's keep it simple for now at first. My settings are similar, only the sole flag I have is rw-- if you remove those flags, does the speed change at all? no, still the same 20-60KBps that hovers about 30KBps. Doing a tcpdump did not reveal much: 04:06:02.160304 IP (tos 0x0, ttl 64, id 23123, offset 0, flags [none], proto: UDP (17), length: 152) client.2131922575 > fserver.nfs: 124 access fh 1107,752457/8932452 003f 04:06:02.160506 IP (tos 0x0, ttl 64, id 15711, offset 0, flags [none], proto: UDP (17), length: 148) fserver.nfs > client.2131922575: reply ok 120 access attr: REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166154965.00 1166039237.00 1137328342.00 c 001f 04:06:02.160540 IP (tos 0x0, ttl 64, id 23124, offset 0, flags [none], proto: UDP (17), length: 152) client.2131922576 > fserver.nfs: 124 access fh 1107,752457/8932452 003f 04:06:02.160660 IP (tos 0x0, ttl 64, id 15712, offset 0, flags [none], proto: UDP (17), length: 148) fserver.nfs > client.2131922576: reply ok 120 access attr: REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166154965.00 1166039237.00 1137328342.00 c 001f 04:06:02.160760 IP (tos 0x0, ttl 64, id 23125, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922577 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 0 04:06:02.160781 IP (tos 0x0, ttl 64, id 23126, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922578 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 8192 04:06:02.161204 IP (tos 0x0, ttl 64, id 15713, offset 0, flags [+], proto: UDP (17), length: 1500) fserver.nfs > client.2131922577: reply ok 1472 read REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166154998.00 1166039237.00 1137328342.00 8192 bytes 04:06:03.180538 IP (tos 0x0, ttl 64, id 23127, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922577 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 0 04:06:03.180985 IP (tos 0x0, ttl 64, id 15715, offset 0, flags [+], proto: UDP (17), length: 1500) fserver.nfs > client.2131922577: reply ok 1472 read REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166154999.00 1166039237.00 1137328342.00 8192 bytes 04:06:04.188613 IP (tos 0x0, ttl 64, id 23128, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922578 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 8192 04:06:04.189048 IP (tos 0x0, ttl 64, id 15716, offset 0, flags [+], proto: UDP (17), length: 1500) fserver.nfs > client.2131922578: reply ok 1472 read REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166155000.00 1166039237.00 1137328342.00 8192 bytes 04:06:07.224834 IP (tos 0x0, ttl 64, id 23129, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922577 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 0 04:06:07.225297 IP (tos 0x0, ttl 64, id 15717, offset 0, flags [+], proto: UDP (17), length: 1500) fserver.nfs > client.2131922577: reply ok 1472 read REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166155003.00 1166039237.00 1137328342.00 8192 bytes 04:06:12.265223 IP (tos 0x0, ttl 64, id 23130, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922578 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 8192 04:06:12.265665 IP (tos 0x0, ttl 64, id 15718, offset 0, flags [+], proto: UDP (17), length: 1500) fserver.nfs > client.2131922578: reply ok 1472 read REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166155008.00 1166039237.00 1137328342.00 8192 bytes 04:06:23.366067 IP (tos 0x0, ttl 64, id 23131, offset 0, flags [none], proto: UDP (17), length: 160) client.2131922577 > fserver.nfs: 132 read fh 1107,752457/8932452 8192 bytes @ 0 04:06:23.366554 IP (tos 0x0, ttl 64, id 15719, offset 0, flags [+], proto: UDP (17), length: 1500) fserver.nfs > client.2131922577: reply ok 1472 read REG 644 ids 1001/0 sz 9083425 nlink 1 rdev 45/35717136 fsid 59 fileid 884c64 a/m/ctime 1166155019.00 1166039237.00 1137328342.00 8192 bytes I even tried TCP NFS and nothing. Here is a dmesg: Copyright (c) 1992-2006 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD is a registered trademark of The FreeBSD Foundation. FreeBSD 6.2-PRERELEASE #0: Thu Dec 14 13:55:53 GMT 2006 [EMAIL PROTECTED]:/data/freebsd/obj/usr/src/sys/GENERIC ACPI APIC Table: Timecounter "i8254" frequency 1193182 Hz quality 0 CPU: AMD Sempron(tm) Processor 2800+ (1607.34-MHz 686-class CPU) Origin = "AuthenticAMD" Id = 0x2
Re: Poor NFS performance after recent update
Kimi Ostro wrote: Hi, On 15/12/06, Jay Chandler <[EMAIL PROTECTED]> wrote: What's the entry for the NFS share in /etc/fstab? fserver:/data /media/datanfs rw,-b,-i,-s,-L,noauto I don't run clients with rpc.statd(8) rpc.lockd(8) I do run statd and lockd, but let's keep it simple for now at first. My settings are similar, only the sole flag I have is rw-- if you remove those flags, does the speed change at all? -- Jay Chandler Network Administrator, Chapman University 714.628.7249 / [EMAIL PROTECTED] Today's Excuse: Processes running slowly due to weak power supply ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
Hi, On 15/12/06, Jay Chandler <[EMAIL PROTECTED]> wrote: What's the entry for the NFS share in /etc/fstab? fserver:/data /media/datanfs rw,-b,-i,-s,-L,noauto I don't run clients with rpc.statd(8) rpc.lockd(8) -- Jay Chandler Network Administrator, Chapman University 714.628.7249 / [EMAIL PROTECTED] Today's Excuse: Processes running slowly due to weak power supply -- Kimi ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Poor NFS performance after recent update
Kimi Ostro wrote: I am have a realy big issue with NFS. I updated my fileserver to -STABLE as of 13th December and suffering poor NFS performance. before I was transferring data at around 6-8MBps now it is 30KBps - yes 30KBps!!. also cause some mounted shares to lock up. I thought it was the nve0 interface and swapped it for an fxp0. no change. checked cables. no change. What can I do to debug this further? I feel as though I could take the binary bits and transfer them quicker myself :( What's the entry for the NFS share in /etc/fstab? -- Jay Chandler Network Administrator, Chapman University 714.628.7249 / [EMAIL PROTECTED] Today's Excuse: Processes running slowly due to weak power supply ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Poor NFS performance after recent update
I am have a realy big issue with NFS. I updated my fileserver to -STABLE as of 13th December and suffering poor NFS performance. before I was transferring data at around 6-8MBps now it is 30KBps - yes 30KBps!!. also cause some mounted shares to lock up. I thought it was the nve0 interface and swapped it for an fxp0. no change. checked cables. no change. What can I do to debug this further? I feel as though I could take the binary bits and transfer them quicker myself :( -- Kimi ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"