toasters

toasters@lists.teaparty.net

2 participants
13529 discussions

RE: Experience with Acopia
by Venkat Appineni 21 May '05

21 May '05

Thanks everyone for replying.. From what I heard and have seen the product at customer locations is its reliable and fast... We were just concerned thinking about scenario where EMC would buy Acopia and then the 2 companies (Netapp and EMC) would end up pointing fingers at each other for any issues with NFS :( Other than that anyone know when Netapp is coming up with their Spinnaker integration... -Venkat -----Original Message----- From: John Stoffel [mailto:john.stoffel@taec.toshiba.com] Sent: Thursday, May 19, 2005 11:15 AM To: Jonathan Cragle Cc: Venkat Appineni; toasters(a)mathworks.com Subject: Re: Experience with Acopia Jonathan> Pros: Jonathan> 1.. Their CEO is fairly well known in the industry. So? Jonathan> 2.. Microsoft Gold Partner So? Since when has that meant anything? Jonathan> 3.. Won an award from Leading Lights (same publishing Jonathan> group that publishes Byte and Switch) Might mean something... I think some of the pros that you ignore or don't mention are: 1. virtualizes your NFS/CIFS storage so you can move/manage/upgrade that backend without user downtime or them even noticing. I like to think of it as VxVM for NFS/CIFS. In my case, we're looking at it strictly from an NFS point of view. 2. Allows you to do HSM style management. Files which aren't accessed in a certain amount of time can be migrated transparently between backend storage pools. 3. It's not trying to pass through the NFS traffic, it's both a server and client and does it's own translation. This is unlike the Rainfinity box which sat in the middle of the stream and played games with what the clients saw. 4. It's an appliance. I don't have to setup an OS and then applications and get it all configured to make it run. It's pretty much plug and play. Just like NetApps. Jonathan> 1.. Expensive: The ARX1000 lists for $45,000; the ARX6000 Jonathan> lists for $150,000. You'll need two of them to provide Jonathan> redundancy, so you'll need to double both of these prices. Yup, this is certainly a con. It's gotten better price wise if you look at how many Acopia volumes they now allow, versus backends per Acopia volume. But noone has ever said that NetApps are cheap, but they are good devices. Jonathan> 2.. Proprietary file system I have no clue what you're talking about here, the Acopia provides an NFS or CIFS mount point for clients to access. It uses CIFS or NFS to access the backend storage. What filesystem do you see here? Jonathan> 3.. In-band solution: depends on who you ask, but this can Jonathan> directly impact your network utilization (Of course, Jonathan> Acopia will cite a report they paid to have done by ESG Jonathan> (Steve Duplessie group) that states "no impact on network, Jonathan> etc. There aren't enough installations (last I heard they Jonathan> had less than 15 actual customers). Of course it's in-band, it's virtualizing your backend storage, so it does have an impact. In our case, we'll never see it since we're space contrained, not performance contrained. If absolute raw performance is so darn important, why are you using NFS in the first place? Have you looked at how much overhead VxVM puts into a system versus just accessing the disk directly? I never see any complaints there, and I suspect (no real numbers) that VxVM puts some measureable overhead into the loop. But the benefits outweigh the costs by a large margin. Also, if I were setting up an Acopia, I would have my clients connecting via the public LAN, and the Acopia and the backend storage all talking over their own dedicated LAN. Network utilization should be a non-issue then. Jonathan> 4.. Used mostly for file migration and then the appliance Jonathan> is ripped out and moved to the next location for migration Jonathan> of files. I'm not aware of anyone using it for global Jonathan> namespace purposes. So why is this a con? It's more of what the customer is choosing to do with the tool, not with what the tool can do. You need to also say why this is a *bad* thing anyway. And as you say, with only 15 customers so far (maybe more by now), just because some or alot of them have used it for migration, does that mean that the one(s) who are using it for virtualization (global namespace purposes) are doing something extraordinary? Maybe I don't understand your reasoning here, I'd appreciate if you could expand on why you think this is a con of the product as opposed to a mis-use by the customer. Jonathan> 5.. Can't leverage the current computer resources one Jonathan> already has, rather you're paying extra for the hardware. Jonathan> You also can't upgrade the hardware as you see fit, rather Jonathan> you'd need to upgrade to a different switch. When was the last time you took an old computer and turned it into a network switch? Jonathan> 6.. Partners with NetApp, but not OEM'd by them. This is a strength in my mind, since you can use the Acopia to make the transition from one vendor's backend solution to another vendor without hassle. Leverage with your next set of price negotiations with NetApp. Jonathan> Have you already evaluated using VFM software to do this? Jonathan> 1.. It's more cost-effective than Acopia Jonathan> 2.. Sold and supported by NetApp Does the VFM software support me using some other vendors backend storage? For example ,the Acopia would let me keep the currently accessed files on the NetApp, but once they aren't used for a while, I could migrate them automatically to cheaper/slower storage. And then I don't need to do backups either of that data, since I know I have backups elsewhere. Jonathan> 3.. Provides support for NetApp and Microsoft snapshots Jonathan> and more. So why is this an advantage? Jonathan> 4.. Software suite of applications that provide Jonathan> policy-based global namespace, migration, archiving, DR, Jonathan> namespace backup/restore, data classfication & reporting. How is this different from the Acopia? Jonathan> 5.. Award-winning (has won three major industry awards in Jonathan> the past couple of months) Bogus, I could care less whether it won an award, but how does it work in the minds of my peers? Jonathan> 6.. Out-of-band solution that won't impact the network So what if you run the DFM servers on a slow box and it can't keep up with the clients and their needs? What happens then? Jonathan> 7.. Standards-based Which ones? Acopia supports NFS and CIFS. Jonathan> 8.. You can actually try it before you buy it versus going Jonathan> through a huge "selection" process to get an Acopia switch Jonathan> in-house to test in your environment. Download is Jonathan> available. This is a possible advantage, in that you can setup an in house test setup without having to bring hardware in house. I will certainly buy that arguement. But it's not an appliance, which means I now have more work to do setting up the server to host the DFM software. And I have to setup, test, document all the failover steps for DFM as well. So why am I not buying an appliance which has all this work done for me already? And of course, I'm speaking for myself, not my company. John John Stoffel - Senior Staff Systems Administrator - System LSI Group Toshiba America Electronic Components, Inc. - http://www.toshiba.com/taec john.stoffel(a)taec.toshiba.com - 508-486-1087

1 0

Another namespace question.
by Alex Bochannek 20 May '05

20 May '05

Since there has been a lot of discussion about namespace virtualization lately, I thought I'd throw in a question. Nightly, an application creates a fairly large number of files, makes a new directory on an R200, and then dumps the data to it. For a number of reasons, the volume this data is on is 5TB in size and it's currently half full (or half empty, depending on your way of looking at things). The R200 is running 6.5.3P4 and has plenty of space for additional volumes. >From the perspective of the application generating the data as well as from the perspective of the users accessing it, the volume is expected to be accessible under one particular name. The obvious question is: What do I do when /vol/foo runs out of space and I need /vol/foo2? Usually, I'd play automounter tricks since the directories under the volumes are all uniquely named, but I'd rather not. This is NFS-only and the clients are all sorts of different UNIXish systems. Any suggestions? Thanks! Alex.

2 2

Experience with Acopia
by Venkat Appineni 19 May '05

19 May '05

We are planning to use Acopia networks ARX switches to provide global name space to bunch of filers. Has anyone used them before ? Any pros and cons ? -Venkat

3 2

RE: Experience with Acopia
by Kennedy, Jeffrey 19 May '05

19 May '05

I don't know about the pro's but the con's you cite are not accurate in most cases. 1) pretty much the case 2) how do you figure? They use CIFS and NFS just like every other network client. They just happen to be the only client to your servers and the only server to you clients if you go all the way with them. But you don't have to go all the way if you choose not to. You can have just certain file systems under Acopia control. 3) They have owned up to a .5% latency insertion which, given their architecture, I can believe. We'll be looking to verify that claim in our testing. 4) Not hardly, at least not in the NFS world. Since this becomes the mount point for all your clients there's no way someone who needed something like this would be able to tell all the clients "stop doing stuff while I remount all the hosts". This is a put-in-place-and-leave-it-forever appliance. 5) Again, how do you figure? This appliance allows you to mount multiple NFS or CIFS shares from existing servers, aggregate them together into a single view, all the while increasing performance by turning the backend into a storage grid. Are you sure you're talking about Acopia? 6) No idea about that one, but I would use NetApp as the backend of this appliance for certain. The basic idea, given a compute farm and NAS shop (NetApp specifically in this case), is to have the Acopia import however many qtrees you see fit to make up a single file system for the world. Then have it present all of those qtrees as a single file system and mount point. Depending on your policies you can have file writes round robin or do latency based file placement. This is the basic concept of a storage grid, spreading storage IO across multiple systems while still presenting a coherent file system and view. Other people accomplish the same idea in different ways (Ibix, Panasas, Isilon, Lustre, etc..). Acopia is unique in that it puts that capability in the network (switch) rather than in a server appliance or software. I'm not saying that's better, just unique. There are other functions the Acopia can do besides the storage grid but everyone has their own requirements and goals. As to VFM for global namespace, fine if you are a CIFS only shop. They're NFS support is sorely lacking and basically a hack of automounter. They have other functions also, like data replication and site failover, but if you're a mixed environment don't bet on them just yet. ~JK ________________________________ From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Jonathan Cragle Sent: Wednesday, May 18, 2005 9:22 PM To: Venkat Appineni; toasters(a)mathworks.com Subject: Re: Experience with Acopia Pros: 1. Their CEO is fairly well known in the industry. 2. Microsoft Gold Partner 3. Won an award from Leading Lights (same publishing group that publishes Byte and Switch) Cons: 1. Expensive: The ARX1000 lists for $45,000; the ARX6000 lists for $150,000. You'll need two of them to provide redundancy, so you'll need to double both of these prices. 2. Proprietary file system 3. In-band solution: depends on who you ask, but this can directly impact your network utilization (Of course, Acopia will cite a report they paid to have done by ESG (Steve Duplessie group) that states "no impact on network, etc. There aren't enough installations (last I heard they had less than 15 actual customers). 4. Used mostly for file migration and then the appliance is ripped out and moved to the next location for migration of files. I'm not aware of anyone using it for global namespace purposes. 5. Can't leverage the current computer resources one already has, rather you're paying extra for the hardware. You also can't upgrade the hardware as you see fit, rather you'd need to upgrade to a different switch. 6. Partners with NetApp, but not OEM'd by them. Have you already evaluated using VFM software to do this? 1. It's more cost-effective than Acopia 2. Sold and supported by NetApp 3. Provides support for NetApp and Microsoft snapshots and more. 4. Software suite of applications that provide policy-based global namespace, migration, archiving, DR, namespace backup/restore, data classfication & reporting. 5. Award-winning (has won three major industry awards in the past couple of months) 6. Out-of-band solution that won't impact the network 7. Standards-based 8. You can actually try it before you buy it versus going through a huge "selection" process to get an Acopia switch in-house to test in your environment. Download is available. ----- Original Message ----- From: Venkat Appineni <mailto:appinenv@yahoo-inc.com> To: toasters(a)mathworks.com Sent: Wednesday, May 18, 2005 4:44 PM Subject: Experience with Acopia We are planning to use Acopia networks ARX switches to provide global name space to bunch of filers. Has anyone used them before ? Any pros and cons ? -Venkat

1 0

RE: ISCSI Boot Questions
by Derek Lai 19 May '05

19 May '05

Sorry I got the model number wrong. It is QLA 4010C. Derek -----Original Message----- From: Jeff Mohler [mailto:jeff.mohler@signasys.com] Sent: Wednesday, May 18, 2005 6:11 PM To: Derek Lai; 'Sto RageC ' Cc: toasters(a)mathworks.com Subject: Re: ISCSI Boot Questions Boot another server..with the same 4004 card, right? On 5/18/05 6:04 PM, "Derek Lai" <Derek.Lai(a)onyxco.com> wrote: > We are testing the Qlogic 4004 card and it works well. It's pretty > easy to setup. You have it point to a lun and then boot off a CD and > install Windows on that new iSCSI drive. > > I personally think it's a great way to get much higer availability > without the added complexities of a full cluster. We have spare servers standby. > With iSCSI booting, if we run into hardware problem, the operator can > shutdown the server, boot another one in place and be up and running > in less than 5 minutes. > > > Derek > > -----Original Message----- > From: Sto Rage(c) [mailto:netbacker@gmail.com] > Sent: Wednesday, May 18, 2005 4:04 PM > To: Jeff Mohler > Cc: toasters(a)mathworks.com > Subject: Re: ISCSI Boot Questions > > not many cards do that. Qlogic has one, but the demo I saw (in late > 2004) was too painful to configure. Its not ready for primetime I think. > -G > > On 5/18/05, Jeff Mohler <jeff.mohler(a)signasys.com> wrote: >> What card is the most supported/liked for booting a diskless >> workstation via iSCSI these days? >> >> And second, how would you go about migrating a disk booting PC, to an >> iSCSI booting one? >> >> Ciao! >> >> >

1 0

RE: ISCSI Boot Questions
by Derek Lai 19 May '05

19 May '05

You can either have just 1 card that you will need to remove from the machine with the hardware problem and place into another server or in our case we just got 2 cards and we can configure 2 separate machines to boot off the same lun. You just need to make sure that they are not trying to do it at the same time. Only one machine should be up on one lun at a time. Derek -----Original Message----- From: Jeff Mohler [mailto:jeff.mohler@signasys.com] Sent: Wednesday, May 18, 2005 6:11 PM To: Derek Lai; 'Sto RageC ' Cc: toasters(a)mathworks.com Subject: Re: ISCSI Boot Questions Boot another server..with the same 4004 card, right? On 5/18/05 6:04 PM, "Derek Lai" <Derek.Lai(a)onyxco.com> wrote: > We are testing the Qlogic 4004 card and it works well. It's pretty > easy to setup. You have it point to a lun and then boot off a CD and > install Windows on that new iSCSI drive. > > I personally think it's a great way to get much higer availability > without the added complexities of a full cluster. We have spare servers standby. > With iSCSI booting, if we run into hardware problem, the operator can > shutdown the server, boot another one in place and be up and running > in less than 5 minutes. > > > Derek > > -----Original Message----- > From: Sto Rage(c) [mailto:netbacker@gmail.com] > Sent: Wednesday, May 18, 2005 4:04 PM > To: Jeff Mohler > Cc: toasters(a)mathworks.com > Subject: Re: ISCSI Boot Questions > > not many cards do that. Qlogic has one, but the demo I saw (in late > 2004) was too painful to configure. Its not ready for primetime I think. > -G > > On 5/18/05, Jeff Mohler <jeff.mohler(a)signasys.com> wrote: >> What card is the most supported/liked for booting a diskless >> workstation via iSCSI these days? >> >> And second, how would you go about migrating a disk booting PC, to an >> iSCSI booting one? >> >> Ciao! >> >> >

1 0

RE: ISCSI Boot Questions
by Derek Lai 19 May '05

19 May '05

We are testing the Qlogic 4004 card and it works well. It's pretty easy to setup. You have it point to a lun and then boot off a CD and install Windows on that new iSCSI drive. I personally think it's a great way to get much higer availability without the added complexities of a full cluster. We have spare servers standby. With iSCSI booting, if we run into hardware problem, the operator can shutdown the server, boot another one in place and be up and running in less than 5 minutes. Derek -----Original Message----- From: Sto Rage(c) [mailto:netbacker@gmail.com] Sent: Wednesday, May 18, 2005 4:04 PM To: Jeff Mohler Cc: toasters(a)mathworks.com Subject: Re: ISCSI Boot Questions not many cards do that. Qlogic has one, but the demo I saw (in late 2004) was too painful to configure. Its not ready for primetime I think. -G On 5/18/05, Jeff Mohler <jeff.mohler(a)signasys.com> wrote: > What card is the most supported/liked for booting a diskless > workstation via iSCSI these days? > > And second, how would you go about migrating a disk booting PC, to an > iSCSI booting one? > > Ciao! > >

2 1

ISCSI Boot Questions
by Jeff Mohler 19 May '05

19 May '05

What card is the most supported/liked for booting a diskless workstation via iSCSI these days? And second, how would you go about migrating a disk booting PC, to an iSCSI booting one? Ciao!

2 1

best backup for netapps
by Constantin Bogomolnyi 17 May '05

17 May '05

Hello, I'm wondering what is the best backup system for saving big netapps we have tried alot of solutions , but we finished by running 3 sets of netapp in snapmirror so A(master) replicates to B and C (slaves) all have now raid_dp enabled and spare around . So we have realtime B and C backups (we may cluster A and B) . But as its quite expensive "backup" i'm looking for some classic backup with tapes for example but i know from the past that tapes can be big pain to manage (specialy as we have around 2.5T of data on A) Anybody have some "tape backup" that realy works ? sub question : anybody have tried to attach a cheap ide raid shelf to scsi port of the netapp and use it as backup ? today its cheap to have 14x300 gig ide disk array with scsi port , so it could be better than tapes if its directly attached to the netapp , nan ? Best regards cb

2 1

RE: best backup for netapps
by Sphar, Mike 17 May '05

17 May '05

Probably the best question is what exactly do you want these backups for? To be able to restore files people mistakenly delete/corrupt? For full-on disaster recovery? For long-term archiving? I think for the first one, various online or near-offline solutions like you've mentioned (ndmp to another server, etc) are probably fine, but then you could probably do just as well by just relying on snapshots and keep it simple. My personal opinion, which others might disagree with, is that for disaster recovery and *especially* for long-term archiving, you really need good offline storage like tapes. Something you can have shipped to a storage facility buried in a mountain somewhere. With enough money you could have multiple netapps around the country/world with snapmirror or somesuch replicating between them, which is great for DR too, but in the end I'm just not comfortable without something offline to fall back on. Imo the best solution is good backup software and a large automated tape library, but it's not going to be cheap. Even just the tapes can be a huge expense. We're still doing backups over NFS here, largely because when we tried doing NDMP a few years ago it just did not seem ready for prime time, we had too many problems. I'm very much interested in trying NDMP again though, as backing up over NFS is not the most efficient solution. I admit though, I've been in this business a long time and I'm a bit old-fashioned. Other people might be more comfortable than I am with relying entirely on hard drives for disaster recovery. -- Michael W. Sphar - IS&T - Lead Systems Administrator SMBU Engineering Support Services, BMC Software -----Original Message----- From: Constantin Bogomolnyi [mailto:const@carpediem.fr] Sent: Tuesday, May 17, 2005 1:32 PM To: Fox, Adam Cc: Constantin Bogomolnyi; toasters(a)mathworks.com Subject: Re: best backup for netapps Yep , i was ready to bet on it , but how do you want us to backup big netapps ? i have around 2T of videos so it will take _a_lot_ of tapes , and time to backup this . I can always setup a linux nfs server and use ndmp to backup it by a dedicated giga card , buts its a lot of complications . I think backup is a real problem with current huge volumes and cheap ide drives are cheaper than tapes (or close to the tapes) . Cb On Tue, May 17, 2005 at 11:07:41AM -0700, Fox, Adam wrote: > The filer probably will not accept and it is certainly not supported to > hook up any other > RAID arrays to a filer. There is no device support for it. > > -- Adam Fox > adamfox(a)netapp.com > > > > -----Original Message----- > From: Constantin Bogomolnyi [mailto:const@carpediem.fr] > Sent: Tuesday, May 17, 2005 1:42 PM > To: toasters(a)mathworks.com > Subject: best backup for netapps > > Hello, > > I'm wondering what is the best backup system for saving big netapps we > have tried alot of solutions , but we finished by running 3 sets of > netapp in snapmirror so A(master) replicates to B and C (slaves) all > have now raid_dp enabled and spare around . So we have realtime B and C > backups (we may cluster A and B) . > > But as its quite expensive "backup" i'm looking for some classic backup > with tapes for example but i know from the past that tapes can be big > pain to manage (specialy as we have around 2.5T of data on A) > > Anybody have some "tape backup" that realy works ? > > sub question : anybody have tried to attach a cheap ide raid shelf to > scsi port of the netapp and use it as backup ? > > today its cheap to have 14x300 gig ide disk array with scsi port , so it > could be better than tapes if its directly attached to the netapp , nan > ? > > > Best regards > cb > >

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