toasters

toasters@lists.teaparty.net

3 participants
13514 discussions

FS: F630, F720, F740 and F760's with 9gb, 18gb, 36gb shelf options
by ScottUCE＠aol.com 05 Dec '00

05 Dec '00

Used Netapp F630 and F7XX systems are available with 9gb, 18gb and 36gb disk shelf options. Please contact us if you have any interest. Scott Fischmann Union Computer Exchange, Inc. 6233 Idylwood Lane Minneapolis, MN 55436 (952) 935-7282 (952) 935-5056 - FAX www.unioncomputer.com

1 0

RE: Strange problem...
by Traitel, Eyal 05 Dec '00

05 Dec '00

Me and CC-ed Mati had it some time ago... HP/UX just read it with no problems at all. It seems some bug with Solaris I think... We never insetigad it more than just deleting the files from the HP/UX machine... Eyal. ---------------------------------------------------------------------- eTraitel - I'm the new eBuzzword around !!! ---------------------------------------------------------------------- Eyal Traitel - Filer Escalation Engineer CNA, MCSE, CSA, NACA Network Appliance BV Holland Office Center Kruisweg 799b 2132 NG, Hoofddorp The Netherlands Office: +31 23 567 9685 Cellular: +31 6 5497 2568 Email: eyal(a)netapp.com ---------------------------------------------------------------------- Get answers NOW! - NetApp On the Web - http://now.netapp.com ---------------------------------------------------------------------- -----Original Message----- From: Chris Lamb [mailto:skeezics@measurecast.com] Sent: Tuesday, December 05, 2000 4:46 AM To: mitch(a)netline.com Cc: toasters(a)mathworks.com Subject: Re: Strange problem... [...] > du: RCS/ParseNVP.pm,v: Value too large for defined data type [snip] > > Has anybody else seen this problem or have any ideas on how/why this > may have occurred? Yup. I called this in as a bug report about four or five months ago. I could probably dig up the case #, but I think it's closed... This filer that first displayed this problem is running both CIFS and NFS, and I thought (at first) that it might be a bug related to using a "mixed" security style. But the problem appeared even on volumes that were not - nor had they ever been - anything but Unix-style security, nor even accessed once by a CIFS client. The odd thing is, when those files would pop up I could go in and delete them via CIFS... > For what it's worth, the NFS client accessing the files is running > Solaris 5.7. I unfortunately didn't pop over to one of my BSD boxes > before the problem worked itself out. I first encountered this using RCS on Solaris 7 as well (both SPARC & x86). I figured it might be some obscure bug related to file locking and the NFS lock manager, but then someone reported that a _core file_ was displaying the same oddball symptoms... The workaround (if you have CIFS licensed) is to try deleting the file via CIFS. If that doesn't work, try again via NFS - you may find that just as mysteriously as it appeared, the error will then go away and you'll be able to delete (or access) the file. Tickling it via CIFS seems to get it "unstuck." Very odd. And, I know, not very scientific. ;-) I suppose I should have shut down the box and wack'ed it upside its WAFL, just to see if there really is some insidious directory corruption waiting to bite me on the backside at some inopportune time. But I don't really have the luxury of an extended downtime, and since it appears there's a crude workaround, I haven't (to my deep shame :-) re-opened the case and pestered for a fix... -- Chris -- Chris Lamb, Unix Guy MeasureCast, Inc. 503-241-1469 x247 <skeezics(a)measurecast.com>

1 0

RE: Backup solution for 2x F840, 2xF630
by Collie, Michael 05 Dec '00

05 Dec '00

Hello, Manny's response addressed some key questions that an administrator must answer when designing a backup solution. Further, one of the benefits of FC is that it provides backup access to storage that is not directly attached to the source data. Therefore, the question of performance on the F630 should not be the only consideration when making you decision. I do not entirely agree with Prem on the recommended approach for 'keeping things simple'. In fact direct attached SCSI can be more expensive and more complicated to manage. ATL is scheduled to ship a P3000 library this month that supports NDMP. This library supports GbE interface for each of its 16 DLT drives. Each DLT drive within the library is independent (own IP address & NDMP server) which almost eliminates the possibility of a complete library disaster. GbE is a proven technology that is considered mature. My recommendation is to centralize you backup using a Library with enough capacity for future scaling. I do not have enough information on your total infrastructure, i.e. FC, SCSI, Ethernet to make a recommendation on what interface this library should use. This information is necessary to calculate you total cost of ownership for a selected solution. Regards, Michael. -----Original Message----- From: Manny Kaiser [mailto:mkaiser@mmcnet.com] Sent: Monday, December 04, 2000 3:42 PM To: 'Oliver Ho' Cc: toasters(a)mathworks.com Subject: Re: Backup solution for 2x F840, 2xF630 Hi, I would go on and comment, that it is good practice to distant your filers from your backup system, thus a fibre solution is preferred. Today management overhead of one large or several small libraries is not too big when using Veritas, thus I don't think this should be a major consideration. An additional consideration is making sure that the fibre at both ends are compatible. I definitely agree that with snapshots, the biggest concern of a backup system is disaster recovery, thus recovery time is an essential consideration - how long can your business be down until you recover a volume, a filer, or all your filers. The number of drives, the speed that they can actually restore data and of course your volume sizes are key factors in determining a backup system. Manny "Traitel, Eyal" wrote: > Personally, I'd go for the large library because of the price, easier > management for you, better scalability and more features probably... > You will have more flexibility as well... > > On number of drives - consider restore times as well... > Eyal. > > ---------------------------------------------------------------------- > eTraitel - I'm the new eBuzzword around !!! > ---------------------------------------------------------------------- > Eyal Traitel - Filer Escalation Engineer > CNA, MCSE, CSA, NACA > > Network Appliance BV > Holland Office Center > Kruisweg 799b > 2132 NG, Hoofddorp > The Netherlands > Office: +31 23 567 9685 > Cellular: +31 6 5497 2568 > Email: eyal(a)netapp.com > ---------------------------------------------------------------------- > Get answers NOW! - NetApp On the Web - http://now.netapp.com > ---------------------------------------------------------------------- > > -----Original Message----- > From: Oliver Ho [mailto:oliver@interq.or.jp] > Sent: Monday, December 04, 2000 10:22 AM > To: toasters(a)mathworks.com > Subject: Backup solution for 2x F840, 2xF630 > > Dear all, > > I have been given a task of designing a Netapps backup solution for my > network. > > We have > 2x F840 > 2x F630 > and various unix servers. > > Approx 1TB of data for the netapps. > > Various Vendors have informed me that the F630 are too slow for > FibreChannel. > > Solution 1 > One VERY HUGE Library with 2x FC for F840 and 2x SCSI for F630 card > mixture for the netapps > 4 Tape drives > FC Switch for F840s > > Solution 2 > Library with 2x FC for F840, two tape drives > FC Switch for F840s > Library with 2x SCSI for F630, two tape drives > > Solution 3 > Three small library with SCSI for each F630 > One Library with 2x FC with two tape drives > FC Switch for F840s. > > The funny thing is Solution 1 is the cheapest, then 2,3. I guess its > because it mean more hardware. > > Can someone give me some pointers. These are all very expensive solutions. > I have heard that some ppl have used Gigabit ethernet and dumped via NDMP. > > We are thinking of using Veritas Netbackup for backups. > I have heard that you can get some kind of SCSI->FC conversion, so I can > connect to whatever FC connected library I may have. > > What would you do in my case? > > TIA > > Regards > Oliver Ho.

1 0

RE: Backup solution for 2x F840, 2xF630
by Traitel, Eyal 04 Dec '00

04 Dec '00

Personally, I'd go for the large library because of the price, easier management for you, better scalability and more features probably... You will have more flexibility as well... On number of drives - consider restore times as well... Eyal. ---------------------------------------------------------------------- eTraitel - I'm the new eBuzzword around !!! ---------------------------------------------------------------------- Eyal Traitel - Filer Escalation Engineer CNA, MCSE, CSA, NACA Network Appliance BV Holland Office Center Kruisweg 799b 2132 NG, Hoofddorp The Netherlands Office: +31 23 567 9685 Cellular: +31 6 5497 2568 Email: eyal(a)netapp.com ---------------------------------------------------------------------- Get answers NOW! - NetApp On the Web - http://now.netapp.com ---------------------------------------------------------------------- -----Original Message----- From: Oliver Ho [mailto:oliver@interq.or.jp] Sent: Monday, December 04, 2000 10:22 AM To: toasters(a)mathworks.com Subject: Backup solution for 2x F840, 2xF630 Dear all, I have been given a task of designing a Netapps backup solution for my network. We have 2x F840 2x F630 and various unix servers. Approx 1TB of data for the netapps. Various Vendors have informed me that the F630 are too slow for FibreChannel. Solution 1 One VERY HUGE Library with 2x FC for F840 and 2x SCSI for F630 card mixture for the netapps 4 Tape drives FC Switch for F840s Solution 2 Library with 2x FC for F840, two tape drives FC Switch for F840s Library with 2x SCSI for F630, two tape drives Solution 3 Three small library with SCSI for each F630 One Library with 2x FC with two tape drives FC Switch for F840s. The funny thing is Solution 1 is the cheapest, then 2,3. I guess its because it mean more hardware. Can someone give me some pointers. These are all very expensive solutions. I have heard that some ppl have used Gigabit ethernet and dumped via NDMP. We are thinking of using Veritas Netbackup for backups. I have heard that you can get some kind of SCSI->FC conversion, so I can connect to whatever FC connected library I may have. What would you do in my case? TIA Regards Oliver Ho.

2 1

Re: Anyone seen this? .snapshot missing from readdir() even for mount point
by Chris Thompson 04 Dec '00

04 Dec '00

On 14 November, I wrote in toasters about ... [...] > The problem is that sometimes ".snapshot" isn't visible to readdir() > *even* in mount point directories. This seems to happen only in > directories with more than one 4K block (it's notable that "ls -f" > shows .snapshot coming after the first block's worth in such > directories, if it shows up at all), and rather rarely even then > (2 out of over 150 user home directories of >4K that I looked at > today), but as long as the directory isn't tinkered with the state > remains the same. > > I've observed this with several OnTap 5.2.x & 5.3.x releases, and > several different Solaris releases as clients. It *could* be an NFS > client implementation problem, I suppose, but I don't think I believe > that. > > Does this sound familiar to anyone? Thanks to Guy Harris at NetApp this has now been nailed down. It turns out to be a long-standing bug, not yet fixed in any released version of DOT. The problem is that if an NFS client does a READDIRPLUS *starting* at what should be the .snapshot entry (which is indeed interposed between the first and second block) then it gets left out of the results returned. The entry in Bugs Online is 33454, although it's a bit sparse on details at the moment! Chris Thompson University of Cambridge Computing Service, Email: cet1(a)ucs.cam.ac.uk New Museums Site, Cambridge CB2 3QG, Phone: +44 1223 334715 United Kingdom.

2 1

RE: Backup solution for 2x F840, 2xF630
by Premanshu Jain 04 Dec '00

04 Dec '00

I would advice to keep things as simple as possible. The Gig-bit and SAN's(FC) are new unproven technologies. Best way to go for direct attached SCSI's , one attached to each filer. One big library, failure will mess all the filers.A distributed architecture is more safe. You may go for an FC upgradable library and wait for FC technology to mature.. prem Premanshu Jain IS Engineering Support Lead Nortel Networks, IP Services Division 2305, Mission College Blvd, Santa Clara, CA 95054 Direct: (408) 565-3573 ESN: 655-3573 eMail: prjain(a)shastanets.com Web: http://www.nortelnetworks.com/ipservices -----Original Message----- From: Oliver Ho [mailto:oliver@interq.or.jp] Sent: Monday, December 04, 2000 1:22 AM To: toasters(a)mathworks.com Subject: Backup solution for 2x F840, 2xF630 Dear all, I have been given a task of designing a Netapps backup solution for my network. We have 2x F840 2x F630 and various unix servers. Approx 1TB of data for the netapps. Various Vendors have informed me that the F630 are too slow for FibreChannel. Solution 1 One VERY HUGE Library with 2x FC for F840 and 2x SCSI for F630 card mixture for the netapps 4 Tape drives FC Switch for F840s Solution 2 Library with 2x FC for F840, two tape drives FC Switch for F840s Library with 2x SCSI for F630, two tape drives Solution 3 Three small library with SCSI for each F630 One Library with 2x FC with two tape drives FC Switch for F840s. The funny thing is Solution 1 is the cheapest, then 2,3. I guess its because it mean more hardware. Can someone give me some pointers. These are all very expensive solutions. I have heard that some ppl have used Gigabit ethernet and dumped via NDMP. We are thinking of using Veritas Netbackup for backups. I have heard that you can get some kind of SCSI->FC conversion, so I can connect to whatever FC connected library I may have. What would you do in my case? TIA Regards Oliver Ho.

1 0

Copying Data
by Morgan, Joshua 04 Dec '00

04 Dec '00

1. is there a way to copy data within the a filer without having to \\filer1\share <file://\\filer1\share> . What we are trying to do is copy data from a snapshot to a shared directory for restores and would like the entire copy process to be on the filer. Thanks, Joshua Joshua Morgan TriVergent Communications Sr. Lvl Technical Engineer PH: (864) 331-7939 Fax: (864) 271-7810 jmorgan(a)trivergent.com ~~~~~~~~~~~~~~~~~~~ http://www.trivergent.com <http://www.trivergent.com/> ===================

1 0

How long does it take ...
by ushwy1 04 Dec '00

04 Dec '00

Hi, I hope one of you who has vast experience could share some info about the following question: How long does it take to WAFL_chekc (WACK) a file system with roughly about 2 millions inodes? Thanks in advance! Steve. __________________________________________________ Do You Yahoo!? Yahoo! Shopping - Thousands of Stores. Millions of Products. http://shopping.yahoo.com/

2 1

How long does it take ...
by ushwy1 04 Dec '00

04 Dec '00

Hi, I hope one of you who has vast experience can provide some insights regarding the following question: How long does it take to WAFL_check a file system that has roughly about 2 millions inodes, assume there is no inconsistency. Thanks in advance. Steve. __________________________________________________ Do You Yahoo!? Yahoo! Shopping - Thousands of Stores. Millions of Products. http://shopping.yahoo.com/

1 0

RE: Snap Manager for Exchange Question
by Mankovsky, Mike 04 Dec '00

04 Dec '00

Joshua, >1. is there a way to specify the name of the directory that gets made when the snapshot is taken? No. >2. if #1 is no then does anyone have a script to call to the newly created snapshot and copy the data somewhere else? Contact Network Appliance Customer Support for some sample scripts and recommended strategies for archiving SnapManager Snapshots to tape. Regards, Mike Mankovsky Network Appliance Inc. -----Original Message----- From: Morgan, Joshua [mailto:JMorgan@trivergent.com] Sent: Monday, December 04, 2000 9:02 AM To: toasters(a)mathworks.com Subject: Snap Manager for Exchange Question 1. is there a way to specify the name of the directory that gets made when the snapshot is taken? 2. if #1 is no then does anyone have a script to call to the newly created snapshot and copy the data somewhere else? 3. if all the above are no what is a good automated backup strategy when using snap manager for exchange? Joshua Morgan TriVergent Communications Sr. Lvl Technical Engineer PH: (864) 331-7939 jmorgan(a)trivergent.com ~~~~~~~~~~~~~~~~~~~ http://www.trivergent.com <http://www.trivergent.com/> ===================

1 0

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