toasters

toasters@lists.teaparty.net

13520 discussions

Re: Cluster Failover
by Michael Rogan 14 Jul '00

14 Jul '00

Hello, If you force a failover the other system can be up in as little as 30 seconds, the more NFS and CIFS exports you have adds a little time. If the system detects a failure it seems to add 15-20 seconds to give the other system a chance to recover before takeover. Of course it also takes 30+ seconds to giveback. Michael Michael J. Rogan BlackBerry Sys Admin Research In Motion www.blackberry.net www.rim.net -----Original Message----- From: Uros Lampret <uros.lampret(a)ourspace.si> To: toasters(a)mathworks.com <toasters(a)mathworks.com> Sent: Fri Jul 14 01:40:23 2000 Subject: Cluster Failover Hi! Does anybody have any experience with clusters. How much down time can I expect with two F740 in cluster. _____________________________ Uros Lampret Our Space d.o.o. Malnarjeva ulica 47 1000 Ljubljana - Slovenia tel. +386 1 2425 372 fax +386 1 5000 959 email: Uros.Lampret(a)ourspace.si internet: www.ourspace.si _____________________________

1 0

RE: filer Olympics
by Stephen C. Woods 14 Jul '00

14 Jul '00

> Now where's the fun in that??? ;-) You should be able to perform > all your filer hardware maintenance with a hunting knife, some chewing > gum, and two feet of piano wire. :) Uhmmmm I think that you might have ment bailing wire, given that Piano wire will probably take a nick out of your knide blade. ----- Stephen C. Woods; UCLA SEASnet; 2567 Boelter hall; LA CA 90095; (310)-825-8614 Finger for public key scw(a)kinross.seas.ucla.edu,Internet mail:scw@SEAS.UCLA.EDU

1 0

F740 filesystem restore throughput?
by Brian Tao 14 Jul '00

14 Jul '00

How fast can I realistically expect an F740 to write out files doing a restore? I have an F740 that is currently doing nothing else, with a 10/100 Mbps Ethernet NIC connected to a Cisco 3548XL switch, and one shelf of 7x36GB drives. I'm dumping from a Sun E420R (4x450-MHz, 50GB Barracuda's on LSILogic Ultra2/LVD adapters) connected to the same switch via 1000base-SX. The command line used is: ufsdump 0bf 128 - /dev/rdsk/c1t0d0s0 | rsh home-tape restore xbfD 128 - /vol/vol0/local The F740 (home-tape) seems to be the bottleneck in this case: # rsh home-tape sysstat 5 CPU NFS CIFS HTTP Net kB/s Disk kB/s Tape kB/s Cache in out read write read write age 100% 0 0 0 6316 104 29 6127 0 0 >60 99% 0 0 0 6396 106 28 6558 0 0 >60 100% 0 0 0 6356 105 24 7764 0 0 >60 100% 0 0 0 6316 104 40 7189 0 0 >60 93% 0 0 0 5983 99 28 6112 0 0 >60 100% 0 0 0 6395 106 28 6465 0 0 >60 99% 0 0 0 6272 104 51 8878 0 0 >60 100% 0 0 0 6506 107 36 6149 0 0 >60 [...] An 8-hour restore window gives enough time to do about 150GB (which, coincidentally, is how much usable space is on 5x36GB data drives ;-)). Any performance tuning tips for doing a fast-as-possible disaster recovery restore like this? I'm going to try streaming data off tape next, instead of kludging it with a ufsdump, but it seems like the Netapp can't go any faster. Is it possible to tell how much of the CPU is used up in rsh overhead, or TCP/IP overhead? Will I gain much if I used NDMP to stream the data to the filer? What about replacing the Fast Ethernet NIC with a Gigabit NIC that can offload packet checksumming from the main CPU? Anyone have numbers? -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

3 4

RE: filer Olympics
by Anissa.Mohler＠netapp.com 14 Jul '00

14 Jul '00

:From: Brian Tao [mailto:taob@risc.org] :On Thu, 13 Jul 2000, Mike Sphar wrote: :> :> Leatherman? Feh. Cordless drill with adjustable torque. : : Now where's the fun in that??? ;-) You should be able to perform :all your filer hardware maintenance with a hunting knife, some chewing :gum, and two feet of piano wire. :) You just purchased that second GigE card for your filer. You drive over to the co-lo to install it and find out that you forgot your tools in the office. You think to yourself.... "Now, what would MacGyver do?" *grin*

1 0

RE: filer Olympics
by Mike Sphar 14 Jul '00

14 Jul '00

Hey, if I wanted to do things the hard way, I'd be running Linux and Samba on a 386. (Not NT with an NFS server though. That's too brutal even for me.) I made my boss pay good money for these filers for a reason: Less work for me. -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Brian Tao [mailto:taob@risc.org] Sent: Thursday, July 13, 2000 4:55 PM To: Mike Sphar Cc: 'toasters(a)mathworks.com' Subject: RE: filer Olympics On Thu, 13 Jul 2000, Mike Sphar wrote: > > Leatherman? Feh. Cordless drill with adjustable torque. Now where's the fun in that??? ;-) You should be able to perform all your filer hardware maintenance with a hunting knife, some chewing gum, and two feet of piano wire. :) -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

1 0

RE: filer Olympics
by jeff.mohler＠netapp.com 14 Jul '00

14 Jul '00

If you use a long screwdriver with a big fat handle on it..you can spin enough mass once a screw is lose, or just started that you can turn it in in about a second with two fingers on the shaft. How about a yankee drill? *grin* -----Original Message----- From: Brian Tao [mailto:taob@risc.org] Sent: Thursday, July 13, 2000 4:55 PM To: Mike Sphar Cc: 'toasters(a)mathworks.com' Subject: RE: filer Olympics On Thu, 13 Jul 2000, Mike Sphar wrote: > > Leatherman? Feh. Cordless drill with adjustable torque. Now where's the fun in that??? ;-) You should be able to perform all your filer hardware maintenance with a hunting knife, some chewing gum, and two feet of piano wire. :) -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

1 0

filer Olympics
by Todd C. Merrill 13 Jul '00

13 Jul '00

;) Had a humorous thought...we've had to replace a lot of motherboards and memory in the last 2 weeks for various reasons, and I've been getting pretty good at it. I was looking back at the logs at the actual downtime for these various tasks. My best times so far have been, where "downtime" is measured from the filer messgages: System shut down with "halt" command. to NetApp Release 5.w.xRyPz boot complete. memory swap: 3 minutes, 57 seconds [0] motherboard swap: 15 minutes, 54 seconds [0] [1] I've broken the "4-minute mile" for the former, but not the "15 minute" mark for the latter. Next time. Or, maybe a contestant on "Fix That Filer" ala "Name That Tune": FilerAdmin1: I can fix that filer in 6 minutes. FilerAdmin2: I can fix that filer in 5 minutes. FilerAdmin3: I can fix that filer in 4 minutes. Wink Martindale: Okay, FilerAdmin3....fix that filer! Okay, time for some breakfast and some serious work now... [0] This includes a 32 second "penalty" for upping 4 links on a vif. [1] Three PCI cards: NVRAM, FCAL, quad ethernet Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

9 10

RE: filer Olympics
by jeff.mohler＠netapp.com 13 Jul '00

13 Jul '00

*grin* Sounds like some of this chest thumping needs backed up with Netapp Certification! http://www.netapp.com/education/ Of course..I just wanna take Todd to dinner..thats all. -----Original Message----- From: Derek Kelly [mailto:derek.kelly@genomecorp.com] Sent: Thursday, July 13, 2000 11:13 AM To: Todd C. Merrill Cc: toasters(a)mathworks.com Subject: Re: filer Olympics .... perhaps first prize should be a job offer from NetApp? ;) Cheers, Derek Kelly Genomecorp On 7/13/00 at 8:24 AM, tmerrill(a)mathworks.com (Todd C. Merrill) wrote: > > ;) > > Had a humorous thought...we've had to replace a lot of motherboards and > memory in the last 2 weeks for various reasons, and I've been getting > pretty good at it. I was looking back at the logs at the actual > downtime for these various tasks. My best times so far have been, where > "downtime" is measured from the filer messgages: > > System shut down with "halt" command. > > to > > NetApp Release 5.w.xRyPz boot complete. > > > memory swap: 3 minutes, 57 seconds [0] > motherboard swap: 15 minutes, 54 seconds [0] [1] > > I've broken the "4-minute mile" for the former, but not the "15 minute" > mark for the latter. Next time. > > > Or, maybe a contestant on "Fix That Filer" ala "Name That Tune": > > FilerAdmin1: I can fix that filer in 6 minutes. > FilerAdmin2: I can fix that filer in 5 minutes. > FilerAdmin3: I can fix that filer in 4 minutes. > Wink Martindale: Okay, FilerAdmin3....fix that filer! > > Okay, time for some breakfast and some serious work now... > > > [0] This includes a 32 second "penalty" for upping 4 links on a vif. > [1] Three PCI cards: NVRAM, FCAL, quad ethernet > > Until next time... > > The Mathworks, Inc. 508-647-7000 x7792 > 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX > tmerrill(a)mathworks.com http://www.mathworks.com > --- >

1 0

RE: Max SCSI cable length
by mds＠gbnet.net 12 Jul '00

12 Jul '00

On Wed 12 Jul, 2000, Jay Orr <orrjl(a)stl.nexen.com> wrote: > speed of max single max HVD length Max LVD length > fastest device > --------------------------------------------------------------- > 5 MHZ (Scsi1) 6M 25M 12M > 10MHZ (S2 Fast) 3M 25M 12M > 20MHZ (Ult. F20) 1.5M 25M 12M > 40MHZ (Ult2 F40) <nr> 12M 12M > Those are closer to the numbers I was expecting people to come up with. > I'm supprised the F7**'s are HVD (didn't investigate it myself). Seems > like HVD is going [slowly] by the wayside... Well tape drives, like CDROM drives, are slow to match the trends that disks and SCSI-host-adapters set. Even now you can usually feed two high-end tape drives (DLT, AIT, Exabyte) off one Fast, Wide SCSI bus. HVD FW SCSI has been around and widely used for, oh, about 5-6 years I reckon, so it's well-proven, reliable, and R&D-amortised technology. There's a hell of a lot of equipment in the world using HVD, at least I've come across a lot in server rooms and places where, until fibre and LVD came along, it was either HVD FW SCSI or something proprietary from IBM or DEC or you couldn't do the job at all. Until fibre-attached tape drives start selling in higher volume than HVD FW SCSI I don't think I'd claim it to be falling by the wayside. >-- End of excerpt from Jay Orr -- -Mark ... an Englishman in London ...

1 0

RE: Max SCSI cable length
by steve.smart＠netapp.com 12 Jul '00

12 Jul '00

If I have got my facts right ... The LVD SCSI standard is listed as being able to support up to 25m. The HVD standard 5m. Right now the F7** series supports HVD and not LVD. So it seems 17ft is the thing. S -----Original Message----- From: Jeff Bryer [mailto:bryer@sfu.ca] Sent: 11 July 2000 23:07 To: toasters(a)mathworks.com Subject: Max SCSI cable length As we get closer to actually purchasing our new tape library, we've started to think about how things will physically fit in our machine room. According to the NetApp manual, it says the max SCSI length for connecting a tape drive is 17ft. Is that the absolute maximum? I've heard people talk about 25metres. I'm inclined to believe the manual but I would much rather believe 25metres. Otherwise we're going to have to get really creative connecting 4 separate machines (2 of which are F740s) to one tape library via 4 separate SCSI buses on the library (don't ask ;). -- Jeff Bryer bryer(a)sfu.ca Systems Administrator (604) 291-4935 Academic Computing, Simon Fraser University

2 1

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