|
marketingman posted:What do you mean dedicating SAN ports to a particular app, and how does that relate to CIFS? marketingman posted:Local logs for troubleshooting what? Access errors? The same logs exist on the NetApp but I can see an applications team being annoyed at not being able to access that easily... not really a show stopper IMO. I guess I don't really see the added value. It's opinion and I'd love to be proved wrong (ok maybe not, but at least shown the compelling arguments). marketingman posted:+1 on the RAID5 lol, that should stay well away from enterprise storage arrays. 
|
#
?
Jan 24, 2012 16:21
|
|
|
|
| # ? Apr 23, 2024 22:14 |
|
|
madsushi posted:Re: CIFS on or off SAN, one big reason is that your NetApp or VMX isn't going to give you the advanced share/file reporting stats that Windows will give you if you run your storage through a Windows server. I like the idea of making a big LUN, deduping it, and then presenting that LUN to Windows and letting it serve out the data. Granted, most customers choose to just toss CIFS on the NetApp and forget about it, but the share reporting features of Windows are one thing to consider.
|
|
#
?
Jan 24, 2012 16:49
|
|
|
Thanks for all the responses, this is super helpful.marketingman posted:So let me stop you right there because there in your scenario there is no performance hit from dedupe or thin provisioning on a NetApp filer. Yeah, I kind-of phrased this wrong. I'm going to have so much free space, I feel I won't need these features (initially, at least), performance hit or not. Great to know, though. marketingman posted:You'd absolutely go NFS and the NetApp would be pretty easy to set up and never ever look at again but I can't compare that to the VNXe so take that as anecdotal... Is NFS harder to setup multipath/high availability? I hope to buy pretty basic/cheap switches for the storage traffic, other than having LACP. I haven't been able to get my hands on an install guide yet...
|
|
#
?
Jan 24, 2012 16:51
|
|
|
FlyingZygote posted:Is NFS harder to setup multipath/high availability? I hope to buy pretty basic/cheap switches for the storage traffic, other than having LACP. I haven't been able to get my hands on an install guide yet... I think there's a new version, too. e: beat. evil_bunnY fucked around with this message at 17:17 on Jan 24, 2012 |
|
#
?
Jan 24, 2012 17:09
|
|
|
There is a new version of TR-3749 available at http://www.netapp.com/us/library/technical-reports/tr-3749.html
|
|
#
?
Jan 24, 2012 17:12
|
|
|
I'm just having a big argument with Peter Bright from Ars Technica on their forums, where he's trying to convince me that on block-based filesystem, doing filesystem-level checksumming on a per-file basis is a good thing (i.e. one checksum for a whole file), and that RAID resiliency depends on how the filesystem clusters/blocks are laid out on the logical disk instead of the actual array doing the proper parity calculations.
|
|
#
?
Jan 24, 2012 17:56
|
|
|
Intraveinous posted:The Supermicro controllers were the easier of the parts to rack. The Disk Enclosures are OEM'd by Xyratex, who make enclosures for a wide array of different vendors out there. Info here (yeah, it's a few years old and things have likely changed) http://www.theregister.co.uk/2009/03/02/xyratex_sff_arrays/. New, Dell-based Compellent boxes are on their way, slated for Summer release.
|
|
#
?
Jan 24, 2012 18:34
|
|
|
Combat Pretzel posted:I'm just having a big argument with Peter Bright from Ars Technica on their forums, where he's trying to convince me that on block-based filesystem, doing filesystem-level checksumming on a per-file basis is a good thing (i.e. one checksum for a whole file), and that RAID resiliency depends on how the filesystem clusters/blocks are laid out on the logical disk instead of the actual array doing the proper parity calculations.
|
|
#
?
Jan 24, 2012 18:50
|
|
|
Well, nothing. What does a filesystem have to do with a RAID array's operation?
|
|
#
?
Jan 24, 2012 19:09
|
|
|
Combat Pretzel posted:Well, nothing. What does a filesystem have to do with a RAID array's operation?
|
|
#
?
Jan 24, 2012 22:45
|
|
|
evil_bunnY posted:Well CIFS would be routed to end users (so I can't use the same ports for NFS to my virtual hosts). I'd rather put all the ethernets to a couple of 10GBE to my hosts, and then plug those into my not-storage network. Am I doing things wrong? From a Netapp perspective the benefits of running CIFS directly on the box are: 1) Array level BLI snapshots with visibility to the customer through the previous versions tab. Snapshots of LUNS aren't nearly as useful. You can get some of the same functionality with VSS. 2) If you have a second filer snapmirror provides and easy path to DR for CIFS data. 3) dedupe will provide space savings directly to the filesystem, rather than simple providing more space in the volume containing the LUN. 4) LUNs require more care and feeding. If the volume containing the LUN fills up the LUN goes offline. If a CIFS share fills up you just get denied write requests. 5) Growing a volume is marginally easier than growing a LUN. Only very slightly easier than with snapdrive though. Obviously not having to maintain and patch a separate server is a big benefit as well. A FAS just provides so much flexibility for working with native WAFL files that I hate not to use it. Snapmirror, flexclone, sisclone, and multi protocol are all pretty awesome.
|
|
#
?
Jan 24, 2012 23:31
|
|
|
szlevi posted:New, Dell-based Compellent boxes are on their way, slated for Summer release. Yeah... I had expected that to happen sooner, to be honest. That said, our TAM said SC 6.0 *will* enable larger read cache via the onboard memory, and that there is a kit to upgrade the controllers from 6GB to 12GB of RAM to take full advantage of that. He said that if we're insistent, we might be able to get SC 6.0 earlier, but currently it's less than 20 systems running it, and all of them are dev/test systems.
|
|
#
?
Jan 24, 2012 23:32
|
|
|
Misogynist posted:The sanity cost of dealing with Oracle's support, as well as whatever the licensing differential is. On the other hand, Oracle is actual Solaris, and Nexenta is Illumos and a GNU userland. Looking at http://www.youtube.com/watch?v=-zRN7XLCRhc (long, ranty but entertaining talk about SunOS history and how Oracle screwed the OpenSolaris pooch), at the end he mentions that there are a lot of fun features in ZFS among others that will never be picked up by Oracle. I haven't touched Solaris since 10, but those features sound pretty cool.
|
|
#
?
Jan 25, 2012 00:03
|
|
|
evil_bunnY posted:1) This is kinda disingenuous: you want the raw capacity to matter somewhat, but do take into account what you'll pay to get the raid level you'll actually use (if you say RAID5 I'm going to lol) and do take into account what dedup/thin provision will get you. I'm looking at his requirements - he doesn't want to use any of those features so it's pretty reasonable for him to ensure the quotes received give him the capacity he requires without taking these features into account. Some vendors do just give assume you'll use all the software and so wont need so many drives. Also, with regards to the second point i'm confused > an honest question - do you believe no one complains about Netapp or Compellent or is it a case of you've just never seen it / don't know where to look?
|
|
#
?
Jan 25, 2012 06:45
|
|
|
evil_bunnY posted:DFSR is one, virus scanning, not dedicating SAN ports to a particular app (or paying for data movers in EMC's case?), and local logs for troubleshooting. Yeah, the AV point is critical for some environments and I just spoke to someone who was looking to move a shitload of file systems onto a Celerra with a shitload of users. Now AV on Netapp / EMC works like this - you have a plugin on a separate server that interacts with the array. AV software on the servers do all he work. These servers have limited grunt so if you have a load of concurrent users you end up with a load of AV servers and likely additional AV software costs. Net result is the AV solution would have gone from a simple bit of software on the file server to a 3 servers (just for AV), additional AV software, additional EMC software and more blades (as their original environment was not scoped for these bajillions of users) and a ton more complexity. End result was that the best course of action was to do nothing but this was a very specific case. Edit: just remembered that recently Netapp introduced integrated AV so you didn't need a separate server. I think it was for only one AV vendor but really great for environments with few concurrent users.
|
|
#
?
Jan 25, 2012 07:03
|
|
|
Intraveinous posted:
I thin the OS is out about now or at least imminent - http://www.compellent.com/About-Us/News-and-Events/Press-Releases/2012/120111-Storage-Center-6.aspx I think, as you said, it's the hardware that needs to come along and that'll be out later this year. With the extended cache it all depends on your environment - some places that have few databases and many 'non-critical' users (with regards to performance - colleges, universities) couldn't care about extending cache but from the results i've seen with PAm cards and Fast cache i'm massively impressed.
|
|
#
?
Jan 25, 2012 07:10
|
|
|
three posted:Why would you 'lol' at RAID5?
|
|
#
?
Jan 25, 2012 11:36
|
|
|
  So according to whatever sample they're pulling from, 6% of RAID 5 volumes should have data loss. I think I've seen dramatically fewer occurrences, but I guess my career isn't that meticulously documented. That's both scary and a drat good number to use to justify extra budget for RAID 6. Of course, that means ~15/16 RAID 5 volumes don't fail in 5 years.
|
|
#
?
Jan 25, 2012 14:05
|
|
|
Yeah, I've both had bad luck and a *lot* of different customers at the same time. As you up the number of drives it goes to poo poo pretty quick. Also I'm pretty sure they're starting with fixed theoretical failure rates, and drives don't really fail like that.
|
|
#
?
Jan 25, 2012 14:39
|
|
|
Yeah, the real-world constraints that lead to a RAID 5 decision aren't considered. When the person with the money is hell-bent on lots of space for low cost and can't be convinced about the risk, that often leads to other shortcuts and corner-cutting that makes the risk even higher.
|
|
#
?
Jan 25, 2012 15:14
|
|
|
Besides these well-documented corruption/bit rot/etc problems of RAID5 there's the RTO issue: with today's huge drives a resync can take days, giving a much bigger window for another drive to fail during the process... And on top of this you have to factor in vendor-induced issues: recently we had a rush of SATA drive failures in our EQL PS6510E, even went through a double-disk failure and only after several calls EQL support admitted that there might be something fishy going on in their latest firmware and we should watch out for the new update, slated for late January-early February that should fix it. szlevi fucked around with this message at 22:40 on Jan 25, 2012 |
|
#
?
Jan 25, 2012 17:08
|
|
|
Hopefully this is an easy one, but focused more on the networking of a SAN. I have inherited an iSCSI SAN with a single VLAN, single IP network. It looks like this: 1. A SAN with 8 active and 8 failover interfaces, all plugged into multiple core switches 2. Four core switches running in a RSTP spanning-tree ring 3. Four edge switches, each has redundant trunks into two of the core switches 4. Hosts are multi-homed into two of these edge switches While this seems like a valid configuration (everything should tolerate a switch failure), it's not ideal, correct? I'm guessing that there needs to be another VLAN with different IP numbering for a second network to mesh all this together properly and allow for better use of the redundant links via MSTP or multipathing? As far as performance goes we're not touching any of the capacity, so making changes for performance aren't currently needed. My goals at this point are to provide stability and avoid downtime.
|
|
#
?
Jan 25, 2012 20:11
|
|
|
This falls somewhat between the consumer and the enterprise as it's an SMB type question. Environment: 4 server SBS 2008 network running on 3 hosts with ESXi. ~20 workstations. I'm looking at picking up a QNAP 859 or 879 (which allows for a 10GBe card) as an upgrade for a synology which I use for a backup target fileshares/and to keep live backup/test copies of VMs (via iSCSI.) with the new box I'm looking at running a couple low utilization VMs via ESX. I'd probably config the disks for RAID 6. Is it worth putting 10GBe in seeing as it would cost a lot too add the cards to the servers and pick up a switch OR would the bandwidth of 6-8 SATA3 drives in RAID-6 not saturate it anyway? Are there other brands i should be considering?
|
|
#
?
Jan 26, 2012 01:17
|
|
|
Bitch Stewie posted:You know in principle I love Nexenta. I don't know why but I'm just wary about thinking about using it in production. Well the fact that OpenSolaris is no more is kind of a wrench in things. Not to say they can't keep this poo poo working well with what they have, but it sort of limits where they can go with it. Plus I've always been a fan of using "big name" stuff in production. There's just more support if you've got a well known, well worn product. Not to say that Nexenta can't be well supported or anything, hell I am new to the SAN world myself, but I like the idea of using an off the shelf vendor solution if the budget allows. I feel the same way about OpenFiler and FreeNAS too, really. I'm sure they work really well but if I actually had to architect a funded project I would probably start with a big name vendor solution. That's just me. I've adopted so many projects where people used some cheap open source thing to get a task done that I had to eventually swap out because of a lack of support or because they didn't plan for upgrades or whatnot (not talking about storage here, but I think the end result could be the same) that I'm just weary of using anything that isn't well-known with a good long term outlook. Of course all that goes out the window if there's no budget. And for something that's not mission critical like my personal VMware lab, I'm more than happy to use something like Nexenta if it fits. If I break something then nothing of value will have been lost other than some of my own personal time. At the same time though, I sort of asked about Nexenta vs Solaris because I have access to that too and if it does the job just as well then perhaps it's worth boning up on a pure Solaris iSCSI target. some kinda jackal fucked around with this message at 01:38 on Jan 26, 2012 |
|
#
?
Jan 26, 2012 01:36
|
|
|
bob arctor posted:This falls somewhat between the consumer and the enterprise as it's an SMB type question. 10gbE would be overkill for that environment.
|
|
#
?
Jan 26, 2012 01:39
|
|
|
BnT posted:Hopefully this is an easy one, but focused more on the networking of a SAN. I have inherited an iSCSI SAN with a single VLAN, single IP network. It looks like this: It's going to depend on what type of storage you've got, each has it's own preferred way of setting up the network. Some just want a single iSCSI Vlan with all the interfaces on it, other prefer 2 or more vlans.
|
|
#
?
Jan 26, 2012 01:39
|
|
|
OOPS: Microsoft Adds Data Deduplication to NTFS in Windows 8
|
|
#
?
Jan 26, 2012 03:13
|
|
|
szlevi posted:OOPS: Microsoft Adds Data Deduplication to NTFS in Windows 8 I'm excited as poo poo for this new filesystem, though. Welcome to the 21st century, Microsoft!
|
|
#
?
Jan 26, 2012 05:32
|
|
|
Welcome to the new generation, Microsoft's filesystem for their unreleased windows 8 server OS that you can't use on client systems and won't function as a boot drive or work with removable drives and may still get cut from the initial release of the server OS.
|
|
#
?
Jan 26, 2012 07:49
|
|
|
what is this posted:Welcome to the new generation, Microsoft's filesystem for their unreleased windows 8 server OS that you can't use on client systems and won't function as a boot drive or work with removable drives and may still get cut from the initial release of the server OS.
|
|
#
?
Jan 26, 2012 13:47
|
|
|
The fact that in this day and age MBR is still commonplace, and required for Win7 boot on most computers, does not give me great confidence. Yes, partition table isn't file system, and the EFI/BIOS thing is closely wrapped up with MBR/GPT, but seriously guys? You couldn't figure something out? Apple switched cleanly from OpenFirmware to EFI without most people noticing any change...
|
|
#
?
Jan 26, 2012 16:27
|
|
|
what is this posted:Apple switched cleanly from OpenFirmware to EFI without most people noticing any change... You can do poo poo like this when you control the entire spectrum of hardware your software runs on, and your customers don't mind upgrading every two releases. But when you're forced to support hardware that is sometimes going on two decades old you have a few less options.
|
|
#
?
Jan 26, 2012 17:14
|
|
|
what is this posted:The fact that in this day and age MBR is still commonplace, and required for Win7 boot on most computers, does not give me great confidence. While I agree with you fully in principle, the fact that Apple was able to make a change isn't really all that telling. Since Apple has complete control over the entirety of the environment on both hardware and software, it should be a lot easier for them to do than someone like MS trying to make sure it will work on every possible configuration out there from umpteen vendors. ReFS is the one that was supposed to be in Windows 2008, then 2008 R2, and now Windows 8 (if it doesn't get cut again), right?
|
|
#
?
Jan 26, 2012 17:18
|
|
|
Less Fat Luke posted:And doesn't support things like sparse files and hard links. I'm just happy Windows finally has a filesystem that doesn't have a 1989 vintage (HPFS was created for OS/2).
|
|
#
?
Jan 26, 2012 17:47
|
|
|
Is the Microsoft iSCSI Initiator literally unable to recognize when an iSCSI target is configured with more than 1 LUN? I've got a target I'm trying to connect to with two configured LUNS (LUN 0 and LUN 1), and this stupid software only seems to see the partition that is configured as LUN0 WTC Microsoft
|
|
#
?
Jan 26, 2012 18:01
|
|
|
Wicaeed posted:Is the Microsoft iSCSI Initiator literally unable to recognize when an iSCSI target is configured with more than 1 LUN? Ah... no? You should see all the LUNs that that host has access to.
|
|
#
?
Jan 26, 2012 18:07
|
|
|
Pretty weird to just default to blaming MS...
|
|
#
?
Jan 26, 2012 18:22
|
|
|
LUN masking is hard, guys.
|
|
#
?
Jan 26, 2012 18:35
|
|
|
Internet Explorer posted:Ah... no? You should see all the LUNs that that host has access to. Ah, I figured it out. I had to disconnect all the current sessions and reconnect, then rescan for disks to see the newly created LUN
|
|
#
?
Jan 26, 2012 18:49
|
|
|
|
| # ? Apr 23, 2024 22:14 |
|
|
Intraveinous posted:While I agree with you fully in principle, the fact that Apple was able to make a change isn't really all that telling. Since Apple has complete control over the entirety of the environment on both hardware and software, it should be a lot easier for them to do than someone like MS trying to make sure it will work on every possible configuration out there from umpteen vendors. Yeah, Apple by default treats its customers like poo poo so they really didn't care what they think. At all. quote:ReFS is the one that was supposed to be in Windows 2008, then 2008 R2, and now Windows 8 (if it doesn't get cut again), right? Nah, it's WAAAAAAY older than that. The story starts in the 90s, with a supposedly information-centered next-gen Microsoft OS, dreamed up by none other than Gates himself, code named Cairo if I remember correctly... and it was to be built on a brand new object-based file system - I remember hearing about a new, next-gen fs in a class back in the 90s that relies on object metadata, supposedly coming in the next NT, version 5 (today known as Windows 2000)... ...which never happened. But the file system idea did stick and it's got a budget and was subsequently named WinFS and promised that even though it won't come in the soon-to-be-released first unified new OS (Windows XP, that is), it will be in the next one, called Longhorn, along with Microsoft Business Framework (MBF) etc: http://blogs.msdn.com/b/theil/archive/2004/05/24/139961.aspx http://weblogs.asp.net/aaguiar/archive/2004/08/28/221881.aspx Then we heard specs and features have been really scaled back... ...it was merged with ObjectSpaces: http://www.alexthissen.nl/blogs/main/archive/2004/05/22/word-is-out-objectspaces-to-be-merged-with-winfs.aspx ...and then rumors about being delayed again, nothing in Longhorn and a lot of denial. Shortly after the denials, of course, eventually came the admission that indeed, nothing will debut in Longhorn: http://www.microsoft.com/presspass/press/2004/Aug04/08-27Target2006PR.mspx Despite a lot of nerd talk about future this was obviously a death sentence for WinFS and MBF - and couple of years everything was scrapped. MBF was gone in as little as one year: http://www.microsoft-watch.com/content/operating_systems/microsoft_scuttles_plans_for_standalone_microsoft_business_framework.html WinFS died in June 2006, very unceremoniously, in a simple blog post: http://blogs.msdn.com/b/winfs/archive/2006/06/23/644706.aspx Whatever useful left was scrapped and worked into SQL Server 2008, that's it. A classic MS-sized fuckup, spanning a decade or more. szlevi fucked around with this message at 20:34 on Jan 26, 2012 |
|
#
?
Jan 26, 2012 19:41
|
|