Tag Info

Hot answers tagged

22

This really depends on your point of view. If I'm an ISV who needs to launch on the tiniest possible budget but I need a crapload of storage, then yes, a brand-name box will be too expensive and the risk/reward of a home-made FreeNAS box would most likely be an acceptable solution. However, if I'm a mega-multi-national corporation with 10,000 users and I ...


10

In our case it comes down to tiers of storage service. This has come about in large part because different needs have different storage requirements. Our ESX environment has Exchange running inside of it, so we need fast, reliable storage. Our desktop-support function just needs lots of it (disk images), with no requirement for speed. The second type doesn't ...


8

Certainly Cisco MDS 95xx FC switches can have 1Gbps and 10Gbps Fibre-Channel-over-Ethernet line cards added to them to convert regular FC traffic onto Data Centre Ethernet which can then be fed into any FCoE/DCE-capable switch, which in turn could have regular 1Gbps and 10Gbps Ethernet ports. That doesn't answer your question however, I'm unaware of any one ...


6

I think $2500 is unrealistic goal. When you talk about high capacity storage you just MUST have some reliability, and you should also include a backup solution in your budget. Recovering 14 TB of data is not cheap under any circumstance. So, I would say a more appropriate budget would be in $10,000 area. I recently built a high capacity storage for my ...


5

I buy storage that works on day one, needs very little management, is fixed without argument in hours and works consistently - regardless of load - for several years. For me that means brand-name FC SAN and NFS/SMB boxes. It all comes down to requirement, for me buying the cheapest parts and hammering them together would be HUGELY expensive to my business, ...


5

As far as my manager is concerned, both myself and the whole of my 'team' of network engineer types are paid far too much to sit there and make homebrew NAS devices when we can buy finished units off the shelf and there are plenty of other things that can't be purchased off the shelf that we could be getting on with instead. And I tend to agree with him.


5

Assuming you stored your volume shadow copies on the same disk that you just moved the 100GB file onto you're probably out of luck. Undelete on NTFS, especially after making changes to the volume, is iffy at best. Shadow copies aren't backup. They're a nice crutch to help you recover files in some situations, but they shouldn't be relied upon as existing ...


5

You can't. Sata is a one-to-one connection. Even in San setups with multiple paths to multiple servers there is always a single point of failure which is dealt with by having multiple arrays. There are multiple ways you can make that storage available to multiple servers eg. a NAS, but you are just moving the single point of failure around.


5

You are missing some of the basic concepts of ZFS. ZFS is meant to be one large pool of disks. After you have established the pool, you are suposed to put file systems on the disks of the pool. All of these file systems then benefit from the pool's underlying RAID layout and other IO characteristics (ZIL et al). Seeing 1000s of file systems reside on a pool ...


4

A potential use-case for ZFS that would see thousands of filesystem mounts would be an organization that uses ZFS for NFS-mounted home directories. Imagine a large company or small university with zpools dedicated to home directory exports. Each user could have a dedicated ZFS filesystem with separate quotas and parameters. If rolling snapshots are ...


4

You can do this with FCoE. I've seen vendors call this "Unified Network" or "Converged Network". Cisco Nexus 5000 HP's Converged Networking page


4

It's doable to do your storage switching and your network switching on the same device. However, it's a lot more expensive and complicated than just keeping it separate, especially for a smaller environment. While the SFP and LC connections are identical when you're looking at an FC port versus when you're looking at an ethernet fiber port, they're ...


4

Evan, Based on your response "the interface on the iMacs and laptops will be gigabit (theoretical)" - then, as Zoredache pointed out - your theoretical max is 120MB/s. It'll be lower, as Chris S said, allowing for normal network overhead. The system is as fast as the slowest NIC. In this instance, that's the gigabit NICs on the workstation. You also ...


4

Typically this kind of question often falls inbetween us and superuser.com but you're building this for a small business so I think it deserves to be here. For a small environment a Mac Mini is just fine, it has no hardware redundancy (PSU/fans etc.) but it'll do just fine. Firstly make sure you go for the server-specific build, it's more expensive but has ...


4

I imagine you're not using a cluster-aware file system, you're probably using something like FAT, NTFS, ext2/3/4, XFS etc. You'll need to use a cluster file system such as OCFS2, Microsoft Cluster-Services enhanced NTFS, Veritas Storage Foundation, VMFS or others. This is a very basic but all too commonly misunderstood problem for junior storage techs to run ...


4

For some background: The reason you typically get a NAS appliance is to get something that will serve up storage and "just work". Of course you need to do an initial configuration, but after that, you pretty much just leave it alone until a drive fails, or there's a critical firmware update. For a Windows file server, you'll need to do regular ...


4

You can drive nails with the butt-end of a screwdriver, but it won't be efficient. Using the wrong tool for the job is fine in a pinch, but isn't a recipe for long-term results. You're creating cache coherency issues in your proposed "solution" (better hope nobody accessed the old NAS directly, hope that non-ECC RAM on the desktop PC doesn't have bit-rot, ...


3

maybe someone had san and nas on her/his mind?


3

You can't do this with a normal file system, as Chopper3 has pointed out. Trying to do this will quickly lead to data corruption. What you should do is either use a cluster-aware file system, or use a replication technology to synchronize two separate volumes, each presented to a single server.


3

If you're deploying an OSX based system, I'd strongly recommend using Timemachine for your backups versus DIY solutions built around rsync. It's built in, surprisingly good at what it does, and all it needs is enough local disk to play with (or remote disk if you invest in a Time Capsule). If you're looking for ways to find cheap storage, this is not the ...


3

Swift is OpenStack's object storage engine, and as of the Bexar release in Feb 2011, it claims an experimental S3-compatible middleware. As OpenStack is getting a lot of attention (Canonical are shifting Ubuntu to it over Eucalyptus this October, for example) it might be worth a look.


3

This is an odd question and while I do agree with Robert's point that this isn't really suited for this site I do buy an AWFUL lot of disks and I never thought I'd see a post where someone said they were surprised how cheap FC ones were :) Well firstly I have to say that FC disks have always been staggeringly expensive compared to anything else. This was ...


3

It sounds to me like you're describing AFS, the most common implementation of which is OpenAFS. The key OpenAFS concepts are described here: http://docs.openafs.org/UserGuide/ch01.html#HDRWQ3. AFS is: Distributed. Filesystem multiple machines, but still using a unified namespace so the distributed nature is transparent to the client machine. Redundant. ...


3

Without knowing the exact hardware you're using, the max you can get through two SAS SFF-8087 is 24Gbps, or 3 GBps; but many controllers-expander combinations will not actually use all 4 channels in the SFF-8087 correctly and you end up getting approximately a single link (0.75GBps). Considering your performance numbers, I would venture a guess that the ...


3

No 'if's, no 'but's - if you have the option of using 10Gbps FCoE and your configuration has proven stable then it's the best and only way to go. It's still quite new but the efficiencies are overwhelming in comparison to iSCSI, and NFS is just plain 'different'. Be aware however that you should be right up to date with ESX/ESXi 4.1U1 for best FCoE ...


3

In the end, I had to move forward with Eucalyptus Walrus. Eventhough I ended up finding some fundamental issues with it, they all had possible work-arounds. Thanks everyone for your valuable input!


3

This is a recommended solution by Apple, actually. It was one under consideration when I was looking for a way to implement company-wide Apple laptop backups. Since thunderbolt is a high-bandwidth interconnect, there are some very interesting external hardware RAID enclosures available for it. You can run any of these behind a Mini (Ideally using OS X ...


2

You could also look at Cloud Foundry vBlob: https://github.com/cloudfoundry/vblob It's a node.js app (with Ruby wrapper for CF integration) that implements a fair chunk of the S3 protocol on top of any file system that the VM can "see".


2

NFS - is file level storage and is the slowest - routable FCoE - best performance but only if you use it locally into a stub network (is not routable) ISCSI - very good performance but adds a bit of complexity - on the flip side is routable


2

If your goal is ease of use, you may want to consider NFS. It has a mediocre performance overhead (-~5% overall throughput, +~20% storage-related CPU) when compared to the FC. Here's a comparison of NFS vs iSCSI vs FC in a 4Gb and 10Gb environment: http://blogs.netapp.com/virtualstorageguy/2010/01/new-vmware-and-netapp-protocol-performance-report.html



Only top voted, non community-wiki answers of a minimum length are eligible