toasters

toasters@lists.teaparty.net

13515 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

Max SCSI cable length
by Jeff Bryer 12 Jul '00

12 Jul '00

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

RE: volume/shelf containment
by Sam Schorr 11 Jul '00

11 Jul '00

Douglas, I think you are confusing "adapter" with "disk shelf". Unless you have 2 FCAL adapters in your filer, and have connected each shelf to its own adapter, you are not spanning adapters and thus not violating what the manual says. I have many filers with their full complement of disk shelves connecting to the 1 FCAL adapter on the filer head. I have RAID groups that span shelves as will happen when disk spare out. There is no performance penalty in this scenario. Sam Schorr Homestead.com ph: (650) 549-3152 fax: (650) 364-7329 sschorr(a)homestead-inc.com http://www.homestead.com Homestead: It's a new world. Make your place. -----Original Message----- From: Douglas Ritschel [mailto:Douglas.Ritschel@fnc.fujitsu.com] Sent: Monday, July 10, 2000 12:20 PM To: Bruce Sterling Woodcock Cc: toasters(a)mathworks.com Subject: Re: volume/shelf containment This is what the book says: "-Write performance will suffer when a RAID group spans more than one FC-AL or SCSI adapter. -Keep all drives assigned to a RAID group on the same adapter whenever posible. -There is an approximate 10% decrease in write performance when the filer attempts to write to a RAID group spanning two adapters. This is due to inherent limitations in the PCI bus." Bruce Sterling Woodcock wrote: > > Unless I misunderstand some of the bus issues involved, the 10% > performance hit would only occur on writes, and since writes are grouped > and it is unlikely that you are writing to the drives ever second, you'll > never > notice it. > > Reads come off individual disks as needed, so it doesn't matter which > controller they are on. (I guess if you needed to read a whole stripe at > once there's a potential hit, but I'm not sure that necessary.) The point > is, with two volumes you're using both controllers anyway. > > Bruce -- Douglas Ritschel Fujitsu Network Communications Two Blue Hill Plaza, Sixth Floor P.O. Box 1609 Pearl River, NY 10965 Ph: 914.731.2101 Fx: 914-731-2011 Douglas.Ritschel(a)fnc.fujitsu.com

6 7

RE: F740 filesystem restore throughput?
by kevin graham 11 Jul '00

11 Jul '00

On Mon, 10 Jul 2000 johnk(a)netapp.com wrote: > It's actually the other way around - packet checksum offloading is a feature > offered by the current (Intel) gigabit cards that the Alteon cards used > previously did not support. The option I was think of was 'udp_lg_dgram.xmit_cksum.offload' (ferreted off of ntap spec benchmark notes).. Obviously I had the wrong impression about what this controlled. What was it? > options cksum_offload.gbeII on Will enable it today, thanks. ..kg..

1 0

Re: volume/shelf containment
by mds＠gbnet.net 11 Jul '00

11 Jul '00

On Mon 10 Jul, 2000, Frank Smith <fsmith(a)hoovers.com> wrote: > Wouldn't it be at least slightly risky to pull all your spares plus one > of your data drives for the duration of a rebuild? Yes. Therefore it's not really recommended. You can of course replace all the spares as soon as the rebuild is suitably underway - which should allow for further rebuilds on other RAID groups if you have a multi-disk failure of that type. If you have a real disk failure in the RAID group you just artificially pulled a disk from, then you start writing the apology emails and calling the ops to bring the tapes back on-site.. Overall, this creates additional work, and as laziness is a virtue, I, personally, wouldn't go pulling disks to neaten up my RAID groups. (similar case: because the filer uses RAID4 not RAID5, and is a pretty smart cookie, you can achieve Great Things(tm) by not having a spare around and pulling the parity disk in each RAID group in a volume... the risk/benefit analysis is an exercise for the reader) > Frank >-- End of excerpt from Frank Smith -- -Mark ... an Englishman in London ...

3 2

Fw: NDMPD
by Ferdie Mazon 11 Jul '00

11 Jul '00

Sorry for the confusion, here is a copy of the rc file. #Regenerated by registry Fri May 26 17:19:40 PDT 2000 #Auto-generated by setup Tue May 2 19:06:59 GMT 2000 hostname srglfs3 ifconfig e0 `hostname`-e0 mediatype 100tx-fd netmask 255.255.0.0 ifconfig e3a `hostname`-e3a mediatype 100tx-fd netmask 255.255.0.0 route add default 10.0.0.160 1 routed on options dns.domainname brandes.com options dns.enable on options nis.enable off savecore exportfs -a nfs on options autosupport.enable on ndmpd on ndmpd status ----- Original Message ----- From: "Jim Ward" <jimw(a)worksta.com> To: <ferdie(a)san.rr.com> Sent: Friday, July 07, 2000 11:33 AM Subject: Re: NDMPD > Ferdie, > > Have you tried "ndmpd on" in lower case? > > Jim > > -- > __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ > | | > | EMail: jimw(a)worksta.com Tel: +1 (603) 672-8600 x232 | > | WWW: http://www.worksta.com Fax: +1 (603) 672-3154 | > | Post: Workstation Solutions, Inc., Five Overlook Drive, Amherst, NH 03031 | > | | > | * Data Backup and Recovery Solutions for Your Peace of Mind * | > | __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ |

7 9

Toasters list turnround sluggishness
by Chris Thompson 11 Jul '00

11 Jul '00

Having been caught out again by asking a question when there was a list message already sent that answered it, I took a look at my mail folders to see why turnround on the list seems to be so sluggish. There doesn't seem to be any appreciable delay in messages getting to mathworks.com, or being processed by Majordomo. It's the time they sit on the queue at smtp.mathworks.com before being sent out that seems to be the problem: it's typically over 90 minutes. A typical (recent) example: | Received: from [144.212.95.101] (helo=smtp.mathworks.com) | by virgo.cus.cam.ac.uk with esmtp (Exim 3.15 #3) | id 13BmNV-00029g-00 | for cet1(a)cus.cam.ac.uk; Mon, 10 Jul 2000 23:50:18 +0100 | Received: (from majordom@localhost) | by smtp.mathworks.com (8.9.3/8.9.3) id RAA04028; | Mon, 10 Jul 2000 17:17:37 -0400 (EDT) That one is a 93 minute delay (22:50:18 GMT - 21:17:37 GMT). I don't think this is poor SMTP connectivity at our end, but if no-one else is seeing these delays then obviously I will investigate that possibility. Otherwise: can anything be done to reduce the turnround delay? Or is it being imposed for our own good... ? :-] 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.

3 3

cannot lockf
by Jim Harm 10 Jul '00

10 Jul '00

I know there was discussion on this before, but I don't recall any resolution. It started with the binmail program could not NFS lock the file. I have run the lockdump -h command and found the processes and the files. The processes are no longer running on the host. The file no longer exists.(I removed it) The locks are still on the filer. How can I get rid of them? and(anyone) Why does the NetApp box keep the lock for processes/files that don't exist? }}}===============>> LLNL James E. Harm (Jim); jharm(a)llnl.gov System Administrator Compaq Clusters (925) 422-4018 Page: 423-7705x57152

2 2

RE: volume/shelf containment
by abond＠netapp.com 10 Jul '00

10 Jul '00

I'd like to put the cat among the pigeons here(!), and ask why you feel it is necessary or desirable to "contain each volume to a shelf of its own". The same situation you have seen here is likely to occur again if another disk fails (it will, eventually). One reason to buy a filer is to minimise administration and maximise uptime, and this process will increase administration as well as introduce unnecessary downtime. One possible explanation is that you need to move some disks later to another filer - in which case identify at that time which disks comprise the volume you wish to move, and then simply move those. As mentioned by others, if you halt the filer, you can move the disks around with no fear whatsoever of confusing the filer - the disk's identity is not related to the slot in which it finds itself, rather to the label on the disk. It is also desirable to minimise the number of "disk fail" operations, following the maxim of "if it ain't broke, don't fix it" :-) Regards, Andrew > -----Original Message----- > From: Fox, Adam > Sent: 10 July 2000 15:55 > To: toasters(a)mathworks.com > Subject: RE: volume/shelf containment > > > There is a simplier way to do this, but it involves halting > the filer. > > You could simply, halt the filer, place the disks where > you want them, then reboot. > > Sorry I didn't mention this obvious way before, but I > tend to think in terms of no downtime. > > -- Adam Fox > NetApp Professional Services, NC > adamfox(a)netapp.com > > > > > -----Original Message----- > > From: Fox, Adam > > Sent: Monday, July 10, 2000 10:38 AM > > To: makin(a)dnrc.bell-labs.com; toasters(a)mathworks.com > > Subject: RE: volume/shelf containment > > > > > > > > > > > -----Original Message----- > > > From: Bhavnesh Makin [mailto:makin@dnrc.bell-labs.com] > > > Sent: Monday, July 10, 2000 10:04 AM > > > To: toasters(a)mathworks.com > > > Subject: volume/shelf containment > > > > > > > > > We have a F740 filer with two 18 gig shelves. Each shelf is > > > dedicated to > > > a volume.We have reserved disk on each shelf as a > hotspare for that > > > particular volume. So in total we have two hot spares. > > > > > > Last week one of the disks failed on the first volume and > > failed disk > > > was recontructed on the spare which was physically > present on second > > > shelf. So we ended up with the volume which transgresses over to > > > another shelf. Our goal was to contain each volume to a > shelf of its > > > own. > > > > > > First,is there a way to contain the volumes with hot > > spares? Secondly > > > how do I make the first volume give up the disk on second > shelf and > > > recontruct it on spare which is physically present on first shelf. > > > > As far as I understand it, spares are global to the filer. > The filer > > will pick the smallest spare disk that will fit. So unless you have > > different sized disks on different volumes, you can't make a > > filer pick > > one disk over another. > > > > As far as fixing your current situation. Config your system > > now so that > > it only have 1 spare (it may be that way anyway now), then > > fail over the > > spare in the 2nd shelf which is now a data drive for the > volume on the > > first shelf. In my opinion, the best way to do this is to > > run the 'disk fail' > > command rather than pull the drive. I think it's cleaner > > since you won't > > have to cause your lop to reset. Then you will force your > > volume to rebuild > > on the only spare in the system, which not so coincidentally > > is the one you > > want. > > > > Hope this helps. > > > > -- Adam Fox > > NetApp Professional Services, NC > > adamfox(a)netapp.com > > >

3 2

RE: volume/shelf containment
by Adam.Fox＠netapp.com 10 Jul '00

10 Jul '00

I 'm new at this NetApp stuff so you may want to get a second opinion but if you pull the rebuilt spare then the first volume will rebuild it on the other spare. Plug it back in and it is now a dirty spare. Chuck This is true, but if you KNOW the disk is good, you can run the disk_zero command (I think you have to rc_toggle_basic first). Then pop the disk out and put it back in, and *poof*, it's a regular spare again! Little tricks that may help you! -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com

1 0

RE: volume/shelf containment
by Chuck Ross 10 Jul '00

10 Jul '00

> We have a F740 filer with two 18 gig shelves. Each shelf is > dedicated to > a volume.We have reserved disk on each shelf as a hotspare for that > particular volume. So in total we have two hot spares. > > Last week one of the disks failed on the first volume and failed disk > was recontructed on the spare which was physically present on second > shelf. So we ended up with the volume which transgresses over to > another shelf. Our goal was to contain each volume to a shelf of its > own. > > First,is there a way to contain the volumes with hot spares? Secondly > how do I make the first volume give up the disk on second shelf and > recontruct it on spare which is physically present on first shelf. > I'm new at this NetApp stuff so you may want to get a second opinion but if you pull the rebuilt spare then the first volume will rebuild it on the other spare. Plug it back in and it is now a dirty spare. Chuck

1 0

RE: volume/shelf containment
by Adam.Fox＠netapp.com 10 Jul '00

10 Jul '00

There is a simplier way to do this, but it involves halting the filer. You could simply, halt the filer, place the disks where you want them, then reboot. Sorry I didn't mention this obvious way before, but I tend to think in terms of no downtime. -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com > -----Original Message----- > From: Fox, Adam > Sent: Monday, July 10, 2000 10:38 AM > To: makin(a)dnrc.bell-labs.com; toasters(a)mathworks.com > Subject: RE: volume/shelf containment > > > > > > -----Original Message----- > > From: Bhavnesh Makin [mailto:makin@dnrc.bell-labs.com] > > Sent: Monday, July 10, 2000 10:04 AM > > To: toasters(a)mathworks.com > > Subject: volume/shelf containment > > > > > > We have a F740 filer with two 18 gig shelves. Each shelf is > > dedicated to > > a volume.We have reserved disk on each shelf as a hotspare for that > > particular volume. So in total we have two hot spares. > > > > Last week one of the disks failed on the first volume and > failed disk > > was recontructed on the spare which was physically present on second > > shelf. So we ended up with the volume which transgresses over to > > another shelf. Our goal was to contain each volume to a shelf of its > > own. > > > > First,is there a way to contain the volumes with hot > spares? Secondly > > how do I make the first volume give up the disk on second shelf and > > recontruct it on spare which is physically present on first shelf. > > As far as I understand it, spares are global to the filer. The filer > will pick the smallest spare disk that will fit. So unless you have > different sized disks on different volumes, you can't make a > filer pick > one disk over another. > > As far as fixing your current situation. Config your system > now so that > it only have 1 spare (it may be that way anyway now), then > fail over the > spare in the 2nd shelf which is now a data drive for the volume on the > first shelf. In my opinion, the best way to do this is to > run the 'disk fail' > command rather than pull the drive. I think it's cleaner > since you won't > have to cause your lop to reset. Then you will force your > volume to rebuild > on the only spare in the system, which not so coincidentally > is the one you > want. > > Hope this helps. > > -- Adam Fox > NetApp Professional Services, NC > adamfox(a)netapp.com >

1 0

RE: volume/shelf containment
by Adam.Fox＠netapp.com 10 Jul '00

10 Jul '00

> -----Original Message----- > From: Bhavnesh Makin [mailto:makin@dnrc.bell-labs.com] > Sent: Monday, July 10, 2000 10:04 AM > To: toasters(a)mathworks.com > Subject: volume/shelf containment > > > We have a F740 filer with two 18 gig shelves. Each shelf is > dedicated to > a volume.We have reserved disk on each shelf as a hotspare for that > particular volume. So in total we have two hot spares. > > Last week one of the disks failed on the first volume and failed disk > was recontructed on the spare which was physically present on second > shelf. So we ended up with the volume which transgresses over to > another shelf. Our goal was to contain each volume to a shelf of its > own. > > First,is there a way to contain the volumes with hot spares? Secondly > how do I make the first volume give up the disk on second shelf and > recontruct it on spare which is physically present on first shelf. As far as I understand it, spares are global to the filer. The filer will pick the smallest spare disk that will fit. So unless you have different sized disks on different volumes, you can't make a filer pick one disk over another. As far as fixing your current situation. Config your system now so that it only have 1 spare (it may be that way anyway now), then fail over the spare in the 2nd shelf which is now a data drive for the volume on the first shelf. In my opinion, the best way to do this is to run the 'disk fail' command rather than pull the drive. I think it's cleaner since you won't have to cause your lop to reset. Then you will force your volume to rebuild on the only spare in the system, which not so coincidentally is the one you want. Hope this helps. -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com

1 0

DLT 7000-----Suggestions
by bhaskar.g＠philips.com 10 Jul '00

10 Jul '00

Hi all, we are planning to purchase a DLT 7000 Tape library. Please can anyone suggest, as to which Library suits the Filer. we are currently having a Filer F720 (preferably HP Make). kind regards Bhaskar G Philips Semiconductors, #3, Sterling Square,Madras Bank Road, Off St Mark's Road,BANGALORE-560 001. Voice : +91 80 222 9200 Ext 5060 Fax: +91 80 2229700

1 0

RE: NDMPD
by Francett, Robert D (SAIC) 07 Jul '00

07 Jul '00

Ferdie, Grasping at straws... After your toaster comes up, if you enter 'ndmpd status' from the console or admin host, is the status 'ON' then? If so, perhaps it's just timing. Maybe the process to switch on ndmpd hasn't completed yet when the 'ndmpd status' command is read. Try moving the 'ndmpd on' up a few lines in the /etc/rc file - like just before the 'options' calls - to give it a bit more time. Bob > ---------- > From: Ferdie Mazon[SMTP:ferdie@san.rr.com] > Sent: Friday, July 07, 2000 10:35 AM > To: toasters(a)mathworks.com > Subject: Fw: NDMPD > > Sorry for the confusion, here is a copy of the rc file. > > #Regenerated by registry Fri May 26 17:19:40 PDT 2000 > #Auto-generated by setup Tue May 2 19:06:59 GMT 2000 > hostname srglfs3 > ifconfig e0 `hostname`-e0 mediatype 100tx-fd netmask 255.255.0.0 > ifconfig e3a `hostname`-e3a mediatype 100tx-fd netmask 255.255.0.0 > route add default 10.0.0.160 1 > routed on > options dns.domainname brandes.com > options dns.enable on > options nis.enable off > savecore > exportfs -a > nfs on > options autosupport.enable on > ndmpd on > ndmpd status > > > ----- Original Message ----- > From: "Jim Ward" <jimw(a)worksta.com> > To: <ferdie(a)san.rr.com> > Sent: Friday, July 07, 2000 11:33 AM > Subject: Re: NDMPD > > > > Ferdie, > > > > Have you tried "ndmpd on" in lower case? > > > > Jim > > > > -- > > __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ > __ > __ > > | > | > > | EMail: jimw(a)worksta.com Tel: +1 (603) 672-8600 x232 > | > > | WWW: http://www.worksta.com Fax: +1 (603) 672-3154 > | > > | Post: Workstation Solutions, Inc., Five Overlook Drive, Amherst, NH > 03031 | > > | > | > > | * Data Backup and Recovery Solutions for Your Peace of Mind * > | > > | __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ __ > __ > __ | >

2 1

RE: NDMPD
by Mike Sphar 07 Jul '00

07 Jul '00

Both of my filers still have the "nfs on" entries and they're working fine, and the entries after that are being executed. -- 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, July 07, 2000 12:25 PM To: Ferdie Mazon; toasters(a)mathworks.com Subject: Re: NDMPD Is the "nfs on" entry even required anymore? Perhaps it's bailing out of the rc file once it reads that line. Bruce

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