Talk:Storage hypervisor
![]() | This article was nominated for deletion on 28 March 2016. The result of the discussion was merge to Software-defined storage. |
![]() | This redirect does not require a rating on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
I have some more info I'm working on. I was wondering if this would be better as a part of the WikiProject Computing area. Any thoughts?
Suggestion on portability
[edit]It is a fact that all well known hypervisors are tied to hardware architectures, be they Intel, SPARC or mainframe. This fact extends to Storage Hypervisors. Interesting that the anonymous TCP/IP address 65.7.182.91 is throwing allegations of industry influence... Johnha (talk) 04:16, 17 October 2012 (UTC)
Greetings all. I suggest that the benefit of a storage hypervisor applies even if the implementation does not depend on portability in the current way it is described (see below):
- The portable nature of the storage hypervisor software refers to its flexibility to run on virtual machine or be hosted on many different platforms in a hardware independent manner similar to virtual server hypervisors.
The features: centrally-managed, generic underlying hardware, one level higher up and all the mentioned benefits are not tied into portability of the software. Looking forward to the feedback! Clausegge (talk) 13:31, 27 April 2012 (UTC)
I agree. Hypervisors of all types usually have intimate hardware connections. This is at the chip level - Intel, vs. Power vs. mainframe. They can be overcome, but it is a major effort and often does not turn out well so rarely are. Most well known hypervisors are tied to hardware architectures. Johnha (talk) 23:53, 30 April 2012 (UTC)
Here is another suggestion: That the storage hypervisor article discusses the different variants. I propose that there are four. As in this piece: Rainmaker Files: The Four Design Options For Storage Hypervisors
Clausegge (talk) 12:58, 1 June 2012 (UTC)
65.7.182.91 (talk) 21:20, 21 July 2012 (UTC)
Interesting that the people who pioneered the space take portability so very seriously and the paid industry analysts are so very eager to allow the hardware and device dependent storage vendors to repeat the history of what happened in storage virtualization. In the late 1990s, when storage virtualization first came out, the pioneers like DataCore Software and Falconstor touted 'portable' software that emulated the key functions of a storage controller in a portable fashion that could run on popular servers versus runnning on vendor specific and dependent server platforms and this empowered users with purchasing power and choice via hardware interchangability. Then the definition was opened up and what happened is the industry analyst who too often are sponsored by the large storage vendors and they worked hard to undermine the original definitions and make it a marketing buzzword and then what we saw was a 'free-for-all' where anyone selling any type of RAID or storage array could claim (and they all did) that they were indeed storage virtualization vendors and talked the talk about software while they sold you a 'hardware box'.
The entry on Wikipedia and the definition were added to Wikipedia last summer and sure enough this summer after the term gets some momentum, the industry analyst community wants to get some press and they are repeating history and trying to do the same undermining of a software definition. The trap is to say the change is minor and that it is no big deal - you just need to add the words into it to say 'it can also do be device dependent', once that is done the trap is set for the 'free-for-all' on the definition and once again the users will get confused talk about hardware in a software virtual world.
It is very clear that the traditional storage hardware and array vendors like IBM, EMC, HDS, HP, etc would like to make sure that a storage hypervisor does not encompass the 'portable' and device independence definition, otherwise they can't define themselves as storage hypervisors. But the term is becoming 'sexy' and if the analysts can turn it into just another marketing 'buzzword' then heck they can write reports on how an array is really a 'storage hypervisor' - OMG! I saw this movie before. Don't undermine the definition.
Let me make it real clear - a storage hypevisor is software. That means you can download it! If they have to ship you hardware it is not a storage hypervisor. If it can't run on a Intel/AMD server, a Dell, HP, IBM, whitebox server or on a VM it is not 'portable' and it is device specific and therefore not a storage hypervisor. If the storage is defined as being the storage that works exclusively within the confines of a 'box' or array that means it is not a storage hypervisor. Bottomline, storage devices will 'come and go' and a storage hypervisor like VMware or Hyper-V are divorced from the device choices which can be SSDs, interanl server disks, any vendor array and even Cloud storage.
65.7.182.91 (talk) 21:20, 21 July 2012 (UTC)
I saw that an anonymous user has reversed updates without engaging in a dialogue on this Talk page. Plea to everyone with an interest in the concept of Storage Hypervisor: Please make a case for changes here. The Help page shows the protocol for making changes.
Clausegge (talk) 08:08, 27 June 2012 (UTC)
The Virtual Machine wikipedia reference was in the article before this recent edit, reverting the edit makes no difference with regard to what http://en.wikipedia.org/wiki/User:Jasper_Deng objected to. Actually is not having wikipedia references in an article +a new requirement? The reference had been in the article for a long time. (and seemed relevant). Fixing that problem, but reverting the most recent edits108.208.235.2 (talk) 05:00, 17 October 2012 (UTC)
- Really? I don't see it in the reverted version.--Jasper Deng (talk) 05:06, 17 October 2012 (UTC)
AfD
[edit]Article nominated for deletion but it turns out it's in the middle of the nowhere so far. If it's not going go be deleted it should be improved: vendor links should be removed and replaced by independent ones. APS (Full Auto) (talk) 19:40, 3 April 2016 (UTC)