toasters March 1999

toasters@lists.teaparty.net

138 participants
192 discussions

Re: 5.2.1 blues
by sirbruce＠ix.netcom.com 04 Mar '99

04 Mar '99

On 03/03/99 08:42:57 you wrote: > >I mean, instead of running 5.1.2P1, P2, D1, D2, D3, D4, D5, D6, or D7 >(deep breath) >I'd like to run 5.1.2 with the patches i feel i'd like to have. I think that's an even worse idea. If you think that Netapp's QA is bad now (and I'm not saying it is), you've just increased their workload exponentially because they have to test each combination of patches to make sure they all work together in whatever combination a user might pick. My experience is that Netapp's patches for specific problems are generally very good, but there always seem to be a couple of large bugs in new releases that many customers run into right away that should have been caught in testing. Bruce

3 2

RE: VIF command
by Mohler, Jeff 04 Mar '99

04 Mar '99

> If you unplug one of the busy links, does one of the idle links pick up > the > slack? > > Sam --- Yes, a spare link picks up the dead one, and holds onto it when you restore the dead link.

1 0

Re: VIF command
by Sam Cramer 04 Mar '99

04 Mar '99

Jeff writes: > I created me a vif interface made up of four 100Mb Ethernet interfaces. > > I see, that only the first two interfaces really do anything, the second two > show no errors, but dont DO anything. It's my understanding that etherchannel switches use a proprietary protocol to distribute load over the links. This protocol often involves a hash of the source and destination addresses. It may be that you have a pattern of access from specific clients that doesn't hash well, and so doesn't distribute the load over all the links. > Im thinking it hasnt reached a 'threshold' to fire up more links. I don't believe that total network load needs to hit a certain threshold in order to use all the links. If you unplug one of the busy links, does one of the idle links pick up the slack? Sam

2 1

RE: NFS 2 -vs-3
by Mohler, Jeff 04 Mar '99

04 Mar '99

Nothing makes me think 2 is -better-, just that I see more "V2 is better" type arguments. Im just ferreting out actual facts *grin* > -----Original Message----- > From: kls(a)netapp.com [SMTP:kls@netapp.com] > Sent: Tuesday, March 02, 1999 10:17 PM > To: jeff.mohler(a)wilcom.com > Cc: toasters(a)mathworks.com > Subject: Re: NFS 2 -vs-3 > > > In a really short list, other than the attribute cache, why is vers 2 > NFS > > better than vers 3? > > What makes you think V2 is better than V3? > > > I will be working with directories with fewer than a dozen LARGE files > on > > each mount. > > With few files per directory, I can't think of anything in favor of > V2, and with large files I can think of various advantages to V3, > larger packet sizes being the major one. > > -- Karl

3 2

Re: Gb-ether and Extreme: closure! :-)
by Bruce Walker 04 Mar '99

04 Mar '99

A very long time ago (mid-November 98), when I got my F720 filer with Gigabit ether, I tried to connect it to an Extreme Summit48, had trouble, contacted NetApp, despite excellent suggestions still had trouble and appealed to this list. Got a few "let us know what happens", but unfortunately no real leads. The story has a happy ending, and here's what did happen. First off: Gigabit on F720 to Summit48 works great. That's settled. Also, with the help I got from NetApp (Mark Smith, Tim McCarthy, David Zwick, Devi Nagaraj, possibly others!), if I had been actually able to concentrate on the problem for more than 30 minutes at a time (and more than 60 minutes per week :-), I might have noticed what was going on earlier. What the problem turned out to be is: on both of the Summit48's I had available to me to test the Gig ether link, the port I was using, always port 50, was missing from any VLAN list. What *that* means is that although the port appeared alive, no packets were ever delivered to that port from any other and it quietly acted as a packet sync for anything received from the filer. Now how did port 50 end up off of all VLANs, on both switches? Simple! Right after I first received my brand new switches, I had trunked the two fiber links on the two Summit48's together, just to see if it would work. I ran fiber from port 49 on one to port 49 on the other and 50 to 50. When the F720 arrived, I moved one fiber link (port 50) from the second Summit48 over to the 720 and unconfigured the trunking. But what happens when you unconfigure a trunked port on a Summit, apparently, is that the port ends up not being on *any* VLAN. The port is "alive" but nothing is routed to or from it. So, I reconfigured the affected ports (put all ports back on the Default VLAN) and the G/B link to the 720 now works just fine. There's still one small "cosmetic" glitch. When the filer comes up, it always reports that the G/B link is up, then right away that it's down. The last console message is always "e8: link down". But it is up (or I couldn't type this mail), so I'm not overly concerned. Do I really need Gigabit ether to an F720? Possibly not. I've only managed to drive the link to 1.5 times the speed of a 100Mbit link so far. But with the prices of Summit48's back when I bought them, I figure the Gig interfaces came for free. Plus, this way I don't have to trunk 100Mbit ether ports and waste good sockets on the switch. And anyway, I'm always adding more clients to load down the filer. That reminds me: has the Progress db been certified yet? -- -bmw | Double helix in the sky tonight | Throw out the hardware | Let's do it right -- Steely Dan; Aja

2 1

newbie asking about autoloaders
by Kathy Fischer 04 Mar '99

04 Mar '99

Hi everyone, I'm brand new to this mail group, and to toasters. We just recently setup our new toaster (720), and we are currently implementing into our environment. Part of this process is developing a backup policy for our toaster, and hooking up an autoloader! Can anyone out there recommend an autoloader (DLT type), or warn against any for use with the toaster? We need to buy one pretty quick, and I want to be sure we get one that won't cause us major headaches :-) !! FYI, we've been looking at the DLT4700, but are unsure which brand to go with. Thanks in advance for any help, Kathy Kathleen G. Fischer Research Systems, Inc. (303)413-3909 kfischer(a)rsinc.com

2 1

RE: Dell Deal
by Strange, Steve 04 Mar '99

04 Mar '99

> From: Dane Jasper [mailto:dane@sonic.net] > > If only the power switch wasn't on the front. We had to tape a half of a > plastic backup tape box on unit to keep someone from pushing the big black > button. You must have an F630 or earlier filer -- your wish came true on the 700-series! Steve

1 0

"5.1.2P2D7"? Whoa, Nelly.
by Brian Rice 03 Mar '99

03 Mar '99

I had no idea that x.y.z.Pm.Dn could exist. So thanks for the tip (sigh). I'm guessing that 5.1.2P2D7 is probably my best bet for a retrograde target, if it comes to that. Any horror stories with that one? I'm not running any cutting-edge hardware...just boring old SCSI shelves (well, one of the F630's is FC). I just want all my filers running the same release, and not panicking when I do a "quota on". It'd be lovely if said release didn't have a name suggestive of a Star Wars droid, but I'll take what I can get. Brian Rice (known to his friends as B2R2)

1 0

Re: Dell Deal
by Jay Sekora 03 Mar '99

03 Mar '99

Dave "Color Me Cool" Hitz wrote: > The most consistent complaint that we've had about the > 700 series is that the bezel isn't "cool" enough. > > I've done some serious rabble rousing, and I've gotten > a strong commitment that the bezel on our next high > end system will be cool. My officemate's response to this was: `They don't think that thing's cool? I think that thing's _mad_ cool! What're they talking about?' So you have at least one voice of approval for the coolness of the 700-series bezels. However, I have to admit that we haven't actually seen the stainless steel ones here. Jay Sekora Northeatern University College of Computer Science PS - I think `mad cool' is supposed to mean something like `very impressive' or `extremely appealing', but I wouldn't know - I don't speak cool.

5 4

RE: Dell Deal
by Bulfer, David 03 Mar '99

03 Mar '99

No, chrome was never really considered. It was supposed to look slightly unrefined, somewhat brash, totally in your face. Aluminum was selected because of the tooling budget that NetApp could afford at the time and because of the expected volume of the F540. Had we known that we'd sell as many F540s as we did, thing may have turned out differently. David -----Original Message----- From: kls(a)netapp.com [mailto:kls@netapp.com] Sent: Wednesday, March 03, 1999 12:20 PM To: hitz(a)netapp.com Cc: jeff.mohler(a)wilcom.com; toasters(a)mathworks.com Subject: Re: Dell Deal > > Is there any way we can get some -real- faces on the F760s? > > > > Seriously, my management was quite unhappy to see four non-matching purple > > plastic faces on our 760s after the sweet Stainless on the demo 630 we had > > *grin* Polished aluminum, actually. It was really supposed to be chrome, but the story I heard was that chrome would have been *really* expensive. > The most consistent complaint that we've had about the > 700 series is that the bezel isn't "cool" enough. > > I've done some serious rabble rousing, and I've gotten > a strong commitment that the bezel on our next high > end system will be cool. A refrigerated bezel? :-) I guess a refrigerated toaster wouldn't make much sense. -- Karl

1 0

← Newer
1
...
13
14
15
16
17
18
19
20
Older →

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 March 1999