toasters

toasters@lists.teaparty.net

13520 discussions

RE: Strange routing problem.
by Mohler, Jeff 28 Jul '99

28 Jul '99

Yes, /etc/netmasks: 10.16.3.0 255.255.255.0 10.16.118.0 255.255.255.0 10.16.119.0 255.255.255.0 10.16.254.0 255.255.255.0 151.142.223.0 255.255.255.0 I didnt know about the /etc/notrouter option...nothin like learning a little everyday! Anyway, rebooting. And... # traceroute 10.16.118.10 traceroute: Warning: ckecksums disabled traceroute: Warning: Multiple interfaces found; using 151.142.223.53 @ hme0 traceroute to 10.16.118.10 (10.16.118.10), 30 hops max, 40 byte packets 1 *^C# Still there. -----Original Message----- From: Tim McCarthy [mailto:tmac@netapp.com] Sent: Monday, July 26, 1999 10:23 AM To: Mohler, Jeff Cc: toasters(a)mathworks.com Subject: Re: Strange routing problem. Did you make a /etc/notrouter file on your solaris box? This prevents the Solaris box from becoming a router at boot time. And how about /etc/netmasks? I would put two entries in just to be safe: 10.16.118.0 255.255.255.0 10.16.119.0 255.255.255.0 --tmac Timothy A. McCarthy --> System Engineer, Eastern Region. Network Appliance http://www.netapp.com 301-230-5847 Office \ / Page Me at:@@ 301-230-5852 Fax \/ 800-654-9619 On Mon, 26 Jul 1999, Mohler, Jeff wrote: > Greetings: > > I have a routing problem on a 2.6 box, here is the detail. > > I want to send traffic out to 10.16.118.20 on qfe0, which will be a private > 10/100 segment to the NetApp for NFS traffic. I am only using 10.16.118.0 > in this message, because 119.0 does the same thing on qfe1. > > This is a prototype box to configure and test first, then deploy to the > field installed Solaris machines. > > The network is like this: > > > NETAPP-F760:e0 public interface connected to 3/1 on a Cat 5505 Switch, on > VLAN 223 (public addressed traffic) > NETAPP-F760:ech1 channel of e1a and e1b on 10.16.118.0 on 5/1-2 > (channel) on VLAN 118 - 10.16.118.10 > NETAPP-F760:ech2 channel of e1c and e1d on 10.16.119.0 on 5/3-4 > (channel) on VLAN 119 - 10.16.119.10 > > On the sun box I have: > > Sun:e0 public interface connected to 3/2 on Cat 5505 on VLAN 223 > Sun:qfe0 10.16.118.23 connected to 3/15 on Cat 5505 on VLAN 118 > Sun:qfe1 10.16.119.23 connected to 3/16 on Cat 5505 on VLAN 119 > Sun:qfe2 10.16.254.23 not yet connected for ADSM traffic > Sun:qfe3 10.16.3.117 connected to management LAN > > > The interface is up, here is netstat output: > > Routing Table: > Destination Gateway Flags Ref Use Interface > -------------------- -------------------- ----- ----- ------ --------- > 151.142.223.0 151.142.223.53 U 2 282 hme0 > 10.16.118.0 10.16.118.23 U 2 5351 qfe0 > 10.16.119.0 10.16.119.23 U 2 0 qfe1 > 10.16.3.0 10.16.3.117 U 2 3 qfe3 > default 151.142.223.1 UG 0 129 > 127.0.0.1 127.0.0.1 UH 0 4 lo0 > > A traceroute to 10.16.118.10 shows the machine choosing hme0 for a > route...why? > > # traceroute 10.16.118.10 > traceroute: Warning: ckecksums disabled > traceroute: Warning: Multiple interfaces found; using 151.142.223.53 @ hme0 > > > I have an NFS mount on the Sun box to the filer using the 10.16.118.10 > address, and it even shows up as the propler hostname:/<path> in the 'df' > output. > > I can blow large amounts of dd traffic to/from the mount, and I see NO > traffic on the correct Channel (ech1) on the Filer. > >

3 2

RE: Non-NDMP backup solutions?
by Melvin, Grant 27 Jul '99

27 Jul '99

Hi, All BudTool and BudTool related products have a release directory on the old Intelliguard ftp site, ftp.iguard.com. They are all accessible via anonymous ftp login. The path to the patches is: pub/dist/patch/<PRODUCT>/<VERSION>/<PATCH#> Cheers, Grant -----Original Message----- From: Jay Orr [mailto:orrjl@stl.nexen.com] Sent: Tuesday, July 27, 1999 12:44 PM To: Rajeev Agrawala Cc: toasters(a)mathworks.com Subject: RE: Non-NDMP backup solutions? Where can budtool patches be obtained from? While we have support and a login for PDC solutions, they don't have the patches... > I am running 5.2.2P1 (recommended release from netapp) on my toaster. > For backups I am using budtool with ndmp for my toaster. I upgraded my > budtool installation from 4.5 to 4.6 with patches 1,3,4,5,10,11,14,17. ----------- Fujitsu - Nexion, St. Louis, MO Jay Orr (314) 579-6517

1 0

(no subject)
by CATULLO＠aol.com 27 Jul '99

27 Jul '99

unsubscribe toasters

1 0

RE: Non-NDMP backup solutions?
by Rajeev Agrawala 27 Jul '99

27 Jul '99

Hi, After reading this thread, it seems that ndmp is the way to go and most of the incompatibilities between various vendors have been taken care off. I am quoting some from parts from this thread below and at the end I am putting my most recent experience with ndmp which was today itself. "Shaun Erickson" wrote: > Because of the frequent upgrades, (which continue to happen, but for > other reasons) we frequently ran into problems because of > incompatibilities between NetApp's implementation of NDMP and BudTool's. > Then, when we'd try to get support, we were always told that the version > of Ontap we were running wasn't certified with BudTool, so it was always > a major pain to get support. To be fair, BudTool broke NDMP on occasion, > as well, with their upgrades. The end result was the same, however. > > "Melvin, Grant" wrote: > I believe this effort has paid off and that we have fixed > our issues in 5.2.*. For 5.3 we ported the NDMP server to > run in the Java Virtual Machine & support V3. Results so > far have shown some bugs, reported on toasters, but these > have been fixed and we have a very stable product. > I am running 5.2.2P1 (recommended release from netapp) on my toaster. For backups I am using budtool with ndmp for my toaster. I upgraded my budtool installation from 4.5 to 4.6 with patches 1,3,4,5,10,11,14,17. The upgrade broke the file history creation/merge part only for my toaster which is using ndmp. For other file systems (Sun/Irix) file history is working fine. I called Intelliguard (now Legato) for support. First thing the support person does is to point me to their certification matrix ( http://www.iguard.com/products/certmatrix1.html ) which does not have any mention of ONTAP 5.2.2 and hence it is not supported. The person tells me that I am out of luck unless I run a supported version. I pointed the person to the piece of code which was broken, and responsible for the problem, the support person agreed to look at that but at the same told me that no change to budtool code can be made for an *Unsupported* ONTAP release which coincidentally is the **Recommended release** from netapp. I guess, that speaks a lot about **COORDINATION**. The choice I have is to either upgrade to an early access relase (5.3.1), which I won't do on a production toaster, or move away from ndmp. rajeev

2 1

Re: ops/second
by alexei＠mindspring.net 27 Jul '99

27 Jul '99

G D Geen <geen(a)msp.sc.ti.com> writes: > officially 7462 NFS ops/sec. We have seen ours run at consistant > ~72K - 74K NFS ops/sec for weeks on end with no complaints from the > user community of performance degredation. 72,000 nfsops? Or did you mean 7200 nfsops? :) Just curious. Alex

3 2

RE: ops/second
by Dawn Hamel 27 Jul '99

27 Jul '99

At SAS, we have seen NFSops/sec in the 25000-30000 range:-) So it definitely depends on what you are doing with the filer Dawn -----Original Message----- From: alexei(a)mindspring.net [mailto:alexei@mindspring.net] Sent: Tuesday, July 27, 1999 12:40 PM To: dmidgett(a)about.com Cc: toasters(a)mathworks.com Subject: Re: ops/second "David Midgett" <dmidgett(a)about.com> writes: > How many ops/second can the F760 handle? It depends on the mix of read/write ops you are doing, as well as what else the filer is busy doing. The f760 is rated (per SPEC, http://www.spec.org/osg/sfs97/results/) is abround 7500 NFSops with the read/write mix they put on it. In one service (with almost equal reads and writes), we see somewhere around 75% cpu with 4k nfsops and 2M/sec in and 2M/sec out, on average. The peaks are much higher, and disk reads/writes are quite high. The cache age is low (3 min). This is on a loaded f760 with 1GB of cache, 3 raid groups of 14 drives (9GB) each. On the other hand, in a mainly read environment I have seen filers push the 9k+ nfsops and pushing 10M/sec with highg cache age, while the cpu is still not higher than 80%. It all depends on your mix and your configuration. Alexei

1 0

Re: ops/second
by alexei＠mindspring.net 27 Jul '99

27 Jul '99

"David Midgett" <dmidgett(a)about.com> writes: > How many ops/second can the F760 handle? It depends on the mix of read/write ops you are doing, as well as what else the filer is busy doing. The f760 is rated (per SPEC, http://www.spec.org/osg/sfs97/results/) is abround 7500 NFSops with the read/write mix they put on it. In one service (with almost equal reads and writes), we see somewhere around 75% cpu with 4k nfsops and 2M/sec in and 2M/sec out, on average. The peaks are much higher, and disk reads/writes are quite high. The cache age is low (3 min). This is on a loaded f760 with 1GB of cache, 3 raid groups of 14 drives (9GB) each. On the other hand, in a mainly read environment I have seen filers push the 9k+ nfsops and pushing 10M/sec with highg cache age, while the cpu is still not higher than 80%. It all depends on your mix and your configuration. Alexei

1 0

Re: ops/second
by Jim Ripley (49er Fan) 27 Jul '99

27 Jul '99

Hi, Checkout the following for this information: http://www.netapp.com/products/level3/f700series_ds.html#F760 Jim -->>From: "David Midgett" <dmidgett(a)about.com> -->>To: <toasters(a)mathworks.com> -->>Subject: ops/second -->>Date: Tue, 27 Jul 1999 11:35:27 -0400 -->>MIME-Version: 1.0 -->>Content-Transfer-Encoding: 7bit -->>X-Priority: 3 (Normal) -->>X-MSMail-Priority: Normal -->>X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 -->>Importance: Normal -->> -->>How many ops/second can the F760 handle? -->>............................................................................ -->>................... -->>David Midgett -->>Manager, Server Operations -->>About.com, Inc. (formerly MiningCo.com, Inc.) -->>http://About.com -->>............................................................................ -->>................. -->> -->> -->> ************************************************************ JIM RIPLEY Escalation Engineer -------------------------- Email: jamesr(a)netapp.com Tech Support Hotline: 1-888-4-NETAPP HQ Address Network Appliance 2770, San Tomas Expressway Santa Clara, CA 95051 bugs, faqs, docs, = http://now.netapp.com/ OS upgrades, etc. ************************************************************

1 0

ops/second
by David Midgett 27 Jul '99

27 Jul '99

How many ops/second can the F760 handle? ............................................................................ ................... David Midgett Manager, Server Operations About.com, Inc. (formerly MiningCo.com, Inc.) http://About.com ............................................................................ .................

2 1

RE: Strange routing problem.
by Mohler, Jeff 26 Jul '99

26 Jul '99

^C# ifconfig -a lo0: flags=849<UP,LOOPBACK,RUNNING,MULTICAST> mtu 8232 inet 127.0.0.1 netmask ff000000 hme0: flags=863<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST> mtu 1500 inet 151.142.223.53 netmask ffffff00 broadcast 151.142.223.255 ether 8:0:20:9f:61:a6 qfe0: flags=863<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST> mtu 1500 inet 10.16.118.23 netmask ffffff00 broadcast 10.16.118.255 ether 8:0:20:9f:61:a7 qfe1: flags=863<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST> mtu 1500 inet 10.16.119.23 netmask ffffff00 broadcast 10.16.119.255 ether 8:0:20:9f:61:a8 qfe2: flags=842<BROADCAST,RUNNING,MULTICAST> mtu 1500 inet 0.0.0.0 netmask 0 ether 8:0:20:9f:61:a9 qfe3: flags=863<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST> mtu 1500 inet 10.16.3.117 netmask ffffff00 broadcast 10.16.3.255 ether 8:0:20:9f:61:a0 -----Original Message----- From: Rajeev Agrawala [mailto:rajeeva@research.bell-labs.com] Sent: Monday, July 26, 1999 3:37 PM To: Mohler, Jeff Cc: Tim McCarthy; toasters(a)mathworks.com Subject: Re: Strange routing problem. What is the output of ifconfig -a on sun machine. Please send that. rajeev "Mohler, Jeff" wrote: > Ok, Ive double checked /etc/netmasks (already was there) /etc/notrouter, > /etc/defaultrouter (already was there) > changed mac addresses on qfe0-qfe3 so the switch doesnt look at all five > interfaces on the machine as one MAC, dbl checked VLAN configs on the Cat > 5505 ports.. > > Still, the Sun wants BADLY to route 10.16.118.0 traffic across the public > hme0 interface, instead of the up/configured/plugged-in 10.16.118.23 > interface. > > Darnit.. > > --- > Did you make a /etc/notrouter file on your solaris box? > This prevents the Solaris box from becoming a router > at boot time. > > And how about /etc/netmasks? > > I would put two entries in just to be safe: > 10.16.118.0 255.255.255.0 > 10.16.119.0 255.255.255.0 > > --tmac > > Timothy A. McCarthy --> System Engineer, Eastern Region. > Network Appliance http://www.netapp.com > 301-230-5847 Office \ / Page Me at:@@ > 301-230-5852 Fax \/ 800-654-9619 > > On Mon, 26 Jul 1999, Mohler, Jeff wrote: > > > Greetings: > > > > I have a routing problem on a 2.6 box, here is the detail. > > > > I want to send traffic out to 10.16.118.20 on qfe0, which will be a > private > > 10/100 segment to the NetApp for NFS traffic. I am only using 10.16.118.0 > > in this message, because 119.0 does the same thing on qfe1. > > > > This is a prototype box to configure and test first, then deploy to the > > field installed Solaris machines. > > > > The network is like this: > > > > > > NETAPP-F760:e0 public interface connected to 3/1 on a Cat 5505 > Switch, on > > VLAN 223 (public addressed traffic) > > NETAPP-F760:ech1 channel of e1a and e1b on 10.16.118.0 on 5/1-2 > > (channel) on VLAN 118 - 10.16.118.10 > > NETAPP-F760:ech2 channel of e1c and e1d on 10.16.119.0 on 5/3-4 > > (channel) on VLAN 119 - 10.16.119.10 > > > > On the sun box I have: > > > > Sun:e0 public interface connected to 3/2 on Cat 5505 on > VLAN 223 > > Sun:qfe0 10.16.118.23 connected to 3/15 on Cat 5505 on VLAN 118 > > Sun:qfe1 10.16.119.23 connected to 3/16 on Cat 5505 on VLAN 119 > > Sun:qfe2 10.16.254.23 not yet connected for ADSM traffic > > Sun:qfe3 10.16.3.117 connected to management LAN > > > > > > The interface is up, here is netstat output: > > > > Routing Table: > > Destination Gateway Flags Ref Use Interface > > -------------------- -------------------- ----- ----- ------ --------- > > 151.142.223.0 151.142.223.53 U 2 282 hme0 > > 10.16.118.0 10.16.118.23 U 2 5351 qfe0 > > 10.16.119.0 10.16.119.23 U 2 0 qfe1 > > 10.16.3.0 10.16.3.117 U 2 3 qfe3 > > default 151.142.223.1 UG 0 129 > > 127.0.0.1 127.0.0.1 UH 0 4 lo0 > > > > A traceroute to 10.16.118.10 shows the machine choosing hme0 for a > > route...why? > > > > # traceroute 10.16.118.10 > > traceroute: Warning: ckecksums disabled > > traceroute: Warning: Multiple interfaces found; using 151.142.223.53 @ > hme0 > > > > > > I have an NFS mount on the Sun box to the filer using the 10.16.118.10 > > address, and it even shows up as the propler hostname:/<path> in the 'df' > > output. > > > > I can blow large amounts of dd traffic to/from the mount, and I see NO > > traffic on the correct Channel (ech1) on the Filer. > > > >

3 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters