toasters

toasters@lists.teaparty.net

13515 discussions

NTFS auditing?
by Brian Tao 07 Dec '00

07 Dec '00

Has anyone heard of "NTFS write auditing"? A group within our company would like that enabled for files in a certain directory so they can track who is modifying files. Is this a feature of NTFS itself, or is this achievable over CIFS to a filer? I'm asking this in the context of a Netapp serving up the data, of course. -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

1 0

Chicago Area Network Appliance
by Knight, Jason 06 Dec '00

06 Dec '00

I am interested in starting a Netapp Users Group in the Chicago Area, I know there are a lot of customers in the suburbs and downtown. I want to gauge the interest before reserving a conference room or making any grand plans. I would like to arrange the meetings close to the majority depending on where the interest is heaviest (assuming enough interest, of course). Please let me know if you are interested and where in ChicagoLand you are located by responding to jasonk(a)wfg.com <mailto:jasonk@wfg.com> . The toasters will probably appreciate NOT being copied on the responses. Thanks, Jason Knight MSCE, Network Appliance CA

1 0

Strange problem...
by mitch＠netline.com 06 Dec '00

06 Dec '00

We just ran across a strange happening on our F760 filer. One of our folks did a "du -s" on an RCS directory to see how much space it was using: => du -s RCS du: RCS/ParseNVP.pm,v: Value too large for defined data type And got a strange error. In fact, pretty much anything that would do a stat(2) call on this file in this directory would generate the same error message. There were no errors logged on the filer and all weekly messages seem to indicate that all is well, but it still puzzles me. I ran off for the requisite caffine boost and came back and am now unable to reproduce the problem. I asked the user if they perhaps did something such as recreating the RCS directory from another source... they hadn't. The file which was complained about is a mere 3422 bytes, and all timestamps indicate that the file has not changed recently. Has anybody else seen this problem or have any ideas on how/why this may have occurred? 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. ~mitch

5 5

RE: Backup solution for 2x F840, 2xF630
by Premanshu Jain 05 Dec '00

05 Dec '00

Hello everybody, I am scheduled to get one P-3000 shortly. If you wish to see my evaluation results, you may wait.. prem -----Original Message----- From: Collie, Michael [mailto:Michael.Collie@netapp.com] Sent: Monday, December 04, 2000 5:40 PM To: 'Manny Kaiser'; 'Oliver Ho'; 'PrJain(a)shastanets.com' Cc: toasters(a)mathworks.com Subject: RE: Backup solution for 2x F840, 2xF630 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: Setting up FTP in NT40 environment
by Benn, Paul 05 Dec '00

05 Dec '00

Hello Paul, Using NT4 and IIS4, you can use a filer on the back end by way of a UNC path but you'll need to configure it for anonymous access. If you want authenticated access, you can do this with IIS5/Win2K and Data ONTAP 5.3.7. The instructions (actually an update to our existing white paper on IIS) aren't yet posted on the web. They should be on the web soon, however. Check this page in a week or so to see the updated information: http://www.netapp.com/tech_library/3078.html <http://www.netapp.com/tech_library/3078.html> Regards, Paul Benn Network Appliance -----Original Message----- From: Maglinger, Paul [mailto:PMAGLINGER@scvl.com] Sent: Tuesday, December 05, 2000 8:43 AM To: 'NetApps list server' Subject: Setting up FTP in NT40 environment We have a F740 running 5.3.6R2 in a WindowsNT environment. We are wanting to FTP to the filer from NT without setting up UIDs and GIDs (basically without anything to do with UNIX). Anyone have any experience with this? Have a prosperous and profitable day! _____ Paul Maglinger, A+, CA, CET, MCP PC Specialist Shoe Carnival Inc. (812)867-4674 pmaglinger(a)scvl.com <mailto:pmaglinger@scvl.com> _____

1 0

Setting up FTP in NT40 environment
by Maglinger, Paul 05 Dec '00

05 Dec '00

We have a F740 running 5.3.6R2 in a WindowsNT environment. We are wanting to FTP to the filer from NT without setting up UIDs and GIDs (basically without anything to do with UNIX). Anyone have any experience with this? Have a prosperous and profitable day! _____ Paul Maglinger, A+, CA, CET, MCP PC Specialist Shoe Carnival Inc. (812)867-4674 pmaglinger(a)scvl.com <mailto:pmaglinger@scvl.com> _____

1 0

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

Snap Manager for Exchange Question
by Morgan, Joshua 04 Dec '00

04 Dec '00

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

Backup solution for 2x F840, 2xF630
by Oliver Ho 04 Dec '00

04 Dec '00

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: Lock files
by Mike Sphar 02 Dec '00

02 Dec '00

Hmmm...delete and recreate the share. Interesting idea. I just may try that. It may not be perfect, but I should be able to automate it at least. (The thing I've been trying to avoid doing was writing a script to identify open files and kill individual locks or sessions.) Thanks for the suggestion. -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Muhlestein, Mark [mailto:mark.muhlestein@netapp.com] Sent: Friday, December 01, 2000 8:18 PM To: 'Mike Sphar'; toasters(a)mathworks.com Subject: RE: Lock files This has been a contentious issue because NT and UNIX locking semantics are so different. We have been considering a number of solutions to this, but in the meantime, one thing some people have done is to have users access data via a share that gets deleted when the build system needs to operate exclusively on a set of files. If you use the "-f" option to "cifs shares -delete" it will unceremoniously kick off all CIFS users of that share. Once the share has been recreated the users can get back into that area. It's not perfect, but it has been workable for some people. It is also possible to run scripts that retrieve information on open files, etc. but I don't know of anyone who has actually done that. Mark -----Original Message----- From: Mike Sphar [mailto:mikey@remedy.com] Sent: Friday, December 01, 2000 7:09 PM To: toasters(a)mathworks.com Subject: RE: Lock files Changing ownership of the file has not had any effect in my experience (this is on NT in general, not just Filers). The denying of access based on an open lock is completely separate from and takes precedence over any ownership as far as I've been able to determine. And no, the Admin generally can't remove a file that has an open lock even if it opened the lock. Try deleting the executable for a process that is currently running and you will probably see what I mean. -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Bruce Sterling Woodcock [mailto:sirbruce@ix.netcom.com] Sent: Friday, December 01, 2000 6:06 PM To: Mike Sphar; toasters(a)mathworks.com Subject: Re: Lock files ----- Original Message ----- From: "Mike Sphar" <mikey(a)Remedy.COM> To: <toasters(a)mathworks.com> Sent: Friday, December 01, 2000 5:53 PM Subject: RE: Lock files > Neither the UNIX root user nor the NT administrator can remove the files in > question when the locks occur. I believe it is a side-effect of the Filer > enforcing NT locks to UNIX clients. (I don't think Netapp is necessarily > doing anything wrong here. I believe this is the correct behavior in most > cases.) I very commonly encounter situations on NT systems where the > administrator cannot remove a file that is open, no matter who owns it. But under NT, can the admin chown (take ownership) of a file that is locked, and then remove it? Also, can the admin remove a file that it itself has locked? > Of course, even if root *could* remove those files as root/administrator, > I'd still like a way for the files owner to be able to remove them. I'd > rather not have to give my build engineers root/administrator access to my > filer. I agree. I believe that UNIX and NT need to respect each-other's permissions, but it should not alter absolute power.. Bruce

1 0

RE: Lock files
by Muhlestein, Mark 02 Dec '00

02 Dec '00

This has been a contentious issue because NT and UNIX locking semantics are so different. We have been considering a number of solutions to this, but in the meantime, one thing some people have done is to have users access data via a share that gets deleted when the build system needs to operate exclusively on a set of files. If you use the "-f" option to "cifs shares -delete" it will unceremoniously kick off all CIFS users of that share. Once the share has been recreated the users can get back into that area. It's not perfect, but it has been workable for some people. It is also possible to run scripts that retrieve information on open files, etc. but I don't know of anyone who has actually done that. Mark -----Original Message----- From: Mike Sphar [mailto:mikey@remedy.com] Sent: Friday, December 01, 2000 7:09 PM To: toasters(a)mathworks.com Subject: RE: Lock files Changing ownership of the file has not had any effect in my experience (this is on NT in general, not just Filers). The denying of access based on an open lock is completely separate from and takes precedence over any ownership as far as I've been able to determine. And no, the Admin generally can't remove a file that has an open lock even if it opened the lock. Try deleting the executable for a process that is currently running and you will probably see what I mean. -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Bruce Sterling Woodcock [mailto:sirbruce@ix.netcom.com] Sent: Friday, December 01, 2000 6:06 PM To: Mike Sphar; toasters(a)mathworks.com Subject: Re: Lock files ----- Original Message ----- From: "Mike Sphar" <mikey(a)Remedy.COM> To: <toasters(a)mathworks.com> Sent: Friday, December 01, 2000 5:53 PM Subject: RE: Lock files > Neither the UNIX root user nor the NT administrator can remove the files in > question when the locks occur. I believe it is a side-effect of the Filer > enforcing NT locks to UNIX clients. (I don't think Netapp is necessarily > doing anything wrong here. I believe this is the correct behavior in most > cases.) I very commonly encounter situations on NT systems where the > administrator cannot remove a file that is open, no matter who owns it. But under NT, can the admin chown (take ownership) of a file that is locked, and then remove it? Also, can the admin remove a file that it itself has locked? > Of course, even if root *could* remove those files as root/administrator, > I'd still like a way for the files owner to be able to remove them. I'd > rather not have to give my build engineers root/administrator access to my > filer. I agree. I believe that UNIX and NT need to respect each-other's permissions, but it should not alter absolute power.. Bruce

1 0

RE: Lock files
by Mike Sphar 02 Dec '00

02 Dec '00

Changing ownership of the file has not had any effect in my experience (this is on NT in general, not just Filers). The denying of access based on an open lock is completely separate from and takes precedence over any ownership as far as I've been able to determine. And no, the Admin generally can't remove a file that has an open lock even if it opened the lock. Try deleting the executable for a process that is currently running and you will probably see what I mean. -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Bruce Sterling Woodcock [mailto:sirbruce@ix.netcom.com] Sent: Friday, December 01, 2000 6:06 PM To: Mike Sphar; toasters(a)mathworks.com Subject: Re: Lock files ----- Original Message ----- From: "Mike Sphar" <mikey(a)Remedy.COM> To: <toasters(a)mathworks.com> Sent: Friday, December 01, 2000 5:53 PM Subject: RE: Lock files > Neither the UNIX root user nor the NT administrator can remove the files in > question when the locks occur. I believe it is a side-effect of the Filer > enforcing NT locks to UNIX clients. (I don't think Netapp is necessarily > doing anything wrong here. I believe this is the correct behavior in most > cases.) I very commonly encounter situations on NT systems where the > administrator cannot remove a file that is open, no matter who owns it. But under NT, can the admin chown (take ownership) of a file that is locked, and then remove it? Also, can the admin remove a file that it itself has locked? > Of course, even if root *could* remove those files as root/administrator, > I'd still like a way for the files owner to be able to remove them. I'd > rather not have to give my build engineers root/administrator access to my > filer. I agree. I believe that UNIX and NT need to respect each-other's permissions, but it should not alter absolute power.. Bruce

1 0

RE: Lock files
by Mike Sphar 02 Dec '00

02 Dec '00

Neither the UNIX root user nor the NT administrator can remove the files in question when the locks occur. I believe it is a side-effect of the Filer enforcing NT locks to UNIX clients. (I don't think Netapp is necessarily doing anything wrong here. I believe this is the correct behavior in most cases.) I very commonly encounter situations on NT systems where the administrator cannot remove a file that is open, no matter who owns it. Of course, even if root *could* remove those files as root/administrator, I'd still like a way for the files owner to be able to remove them. I'd rather not have to give my build engineers root/administrator access to my filer. -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Bruce Sterling Woodcock [mailto:sirbruce@ix.netcom.com] Sent: Friday, December 01, 2000 5:43 PM To: Mike Sphar; toasters(a)mathworks.com Subject: Re: Lock files ----- Original Message ----- From: "Mike Sphar" <mikey(a)Remedy.COM> To: <toasters(a)mathworks.com> Sent: Friday, December 01, 2000 4:11 PM Subject: RE: Lock files > I was about to ask a similar question actually. We keep running into > problems where our build group can't delete a directory because someone else > had opened one of the files in it, and then the build automation fails. Not > even root can remove those files until the offending session is found and > killed or the lock is removed. This I don't understand. Root should always be allowed to remove files even if they are locked. If the filer doesn't allow that, I consider it a bug (even if it's a design choice). Of course, NT's Administrator has different privelages. I suppose you could enforce that on the UNIX side and require root to first chown the files to root, and then delete them. But that would be an optional behavior. Bruce

2 1

RE: Lock files
by Mike Sphar 02 Dec '00

02 Dec '00

I was about to ask a similar question actually. We keep running into problems where our build group can't delete a directory because someone else had opened one of the files in it, and then the build automation fails. Not even root can remove those files until the offending session is found and killed or the lock is removed. Is there a way to selectively disable locking so that some directories/shares/qtrees/etc will allow open files to be deleted? -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: Shaw, Kevin [mailto:Kevin.Shaw@CAX.USA.XEROX.com] Sent: Friday, December 01, 2000 2:02 PM To: 'toasters(a)mathworks.com' Subject: Lock files I'm having problems moving files (using cpio or tar) with the lock bit set. Any ideas/help is appreciated. Thanks in advance. -Kevin

2 1

Lock files
by Shaw, Kevin 02 Dec '00

02 Dec '00

I'm having problems moving files (using cpio or tar) with the lock bit set. Any ideas/help is appreciated. Thanks in advance. -Kevin

2 1

RE: MRTG with netcache.
by Doherty, Niall 02 Dec '00

02 Dec '00

> The goop on now.netapp.com is old, and rather clunky. Yup. I completely agree ;-) > Anybody got any good Netcache MRTG stuff that sucks down all the good > stuff w/o firing up wget to grep out data? Not MRTG, but here are my config files for Cricket (which is sort-of-a-successor to MRTG). If you want to use this with Cricket you should create a subdirectory and dump in the Defaults file and the appropriate "nc" file. NetCache 5.0 has a LOT of info available via SNMP. I don't display *everything* in the "Targets" specification for NetCache-50 but if you read through the files you can see what else might be of interest... Even if you don't use Cricket these files should be of use if you want to use SNMP - the hard part of SNMP is not retrieving the stats, but interpreting and graphing them correctly... Have fun... If you've any questions let me know... Cheers, Niall -- Niall Doherty | Network Appliance Inc. | NetCache Technical Coordinator | 495 East Java Drive | Intercontinental Operations | Sunnyvale, CA 94089, USA | mailto:ndoherty@netapp.com | http://www.netapp.com | Direct Line: +1 408 822 8554 | HQ Phone: +1 408 822 6000 | FAX(->email): +1 801 912 5212 | HQ FAX: +1 408 822 6001 |

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