Posts by _Absinthe_ 14 publicly visible posts • joined 24 Aug 2010
Our list price on the 15TB 2.5" SSD is around the £39k mark, and the 7.5TB is about £20k, so I expect this 30TB monster would be about £80k list price. Of course, those are list prices, what people actually pay for them is obviously much less, perhaps about 40-50% less usually
Quite a few of my customers at work have a decent number of overseas locations, but each location only has a handful of users, so having onsite IT resource at any of them doesn't make sense. I have one customer with about 300 overseas users I think it was, but they're split across about 25+ sites; they only have a (part-time) local IT presence on 2 of those sites as the user count at the rest simply doesn't justify it. A somewhat strange business model which I'll admit I don't understand, but that's probably why I'm the techie, not the commercial guy in the equation :)
There are server vendors who produce servers which have a lot more than 4 CPU sockets. But that's not the main reason the support is there for such a high number of logical cores. The main reason is that Server2016 will likely be around for a good number of years (5-7yrs?), and in that time CPUs will advance quite considerably, like they have in the last 5yrs or so.
From my admittedly rather poor memory, 5yrs ago we were dealing with CPUs which typically had 4,6 or 8 cores, now we're up to 24. In another 5yrs that could quite feasibly be up to ~96. But even if they only progress as far as 48 cores in that time, in a 4-socket box this still means you have 192 physical cores, or 384 logical. All of a sudden that 512 maximum doesn't seem so ludicrous.
FYI - SuperMicro do an 8-way box which I'm told isn't stupidly expensive considering what you can do with it. Loaded up with the top CPU, it gives you 384 logical cores, today.
"outdated networking architectures (no software Defined Networking)" - Then why did Gartner recently call them the worlds most innovative, disruptive and visionary networking vendor? The push they've been making towards open- and software-defined architectures may not have been advertised very well, but they are substantial and very real.
Sure we do.
http://www.theregister.co.uk/2012/06/18/dell_hermes/
The difference is that rather than develop a whole new platform from the ground up to replace a burgeoning legacy architecture, they appear to have simply bought a very disruptive technology (RNA Networks) and are in the process of integrating the tech into another two very disruptive platforms (Equallogic & Compellent) in order to virtualise flash (instead of RAM) into a coherent tier of storage 'outside' of the traditional disk arrays. Basically, what this article is talking about with regards to building 'interfaces' between AFA's and traditional disk arrays, but without needing to begin with two whole platforms and then work backwards to integrate them.
Of course, what will be interesting will be whether they can actually pull it off.
"The El Reg storage desk asked ScaleIO if it could say where the HP StoreVirtual VSA, the Dell EqualLogic VSA and other specific VSAs fall short" There's no such thing as an Equallogic VSA! There is a 'virtual array' sales and training tool, but it has significant limitations in order to prevent it being run in a production environment, and is available internally at Dell only, or to Tier 1 partners with correct authorisation.
"The server has five 2.5-ich drive bays... ...you can't expect more than that in a 2U rack form factor"
The Dell R820, has SIXTEEN 2.5" drive bays!? That is also a 2U 4-socket box with otherwise very similar specs to this DL560 Gen8 (well, the Dell's got an extra PCIe slot, but hey ho).
If we're looking at the Dell R720xd however, which is a 2U, 2-socket box, that one has TWENTY-SIX 2.5" drive bays, and they're ALL hot-pluggable.
I won't even start on the rest of the article.
Guess they must really be losing market share to consider a drastic move like this!? _Giving away_ 3PAR, XP & Lefthand?? I guess if you can't compete technically then price is all you have left. Perhaps a sign that Dells aggressive storage program is really starting to show dividends if HP are this worried?
"An asteroid collision with Earth could now be less likely thanks to a software developer who created a computer program capable of tracking NEOs (Near Earth Objects)."
Without wanting to be too much of a party pooper, the chance of a collision is in no way reduced by some chap writing a piece of software!! We'd simply know about the collision in advance whereas we would previously have been oblivious to the impending doom until we saw the approaching fireball. However, if we actually had a means of destroying the huge rock flying towards us, then it'd be a different story...
(Of course, if I'm wrong, and simply writing a piece of software can alter the trajectories of asteroids just imagine what else we could achieve!!!! :) )
Alien icon because it's obviously the insectoids that are flinging these asteroids at us in the first place! grr, darn those pesky insectoids.
I'm tired due to have a young baby in the house, and am also perhaps missing something pretty major here, but it appears to me that instead of finding a market need, then developing a product to fill it. They appear to have developed a product and only then tried to think of what could be done with it.
The "advanced" features of VMware - vMotion, HA, DRS et al - have been available without needing a dedicated physical SAN for several years now, with LeftHand Networks' VSA for example, and now with VMware's own vSphere Storage Appliance functionality. Not to mention that there are numerous other benefits & reasons why you'd want a dedicated physical SAN other than just to enable those few features I mentioned there, for example, can this product support the VAAI features - Atomic-test-and-set, Data Copy offload, Thin Provision Stun etc etc? Also, done correctly with the right technology - Compellent for example - a dedicated physical SAN with genuinely advanced functionality and a bucket-load of benefits needn't cost the earth either.
That's not to mention that the product this article is advertising doesn't even support ESXi!?!?! What's THAT all about!?!? ESXi has been around for several years now, and the writing has been on the wall pointing towards the obsolescence of full-fat ESX for just as long, so not supporting ESXi is inexcusable. They are basically saying that they cannot run the current version of the market leading virtualisation platform (vSphere 5) and there is no firm date for when they will be able to either, from what I can find.
In addition, this product does away with all the CapEx and OpEx benefits of using commodity servers for your compute nodes, since they have to 'tune & optimise' each node to be able to handle the storage load as well...
All in all, I don't think it's worth the pixels it's being advertised with, but you probably got that already, if indeed you've even read this far :)
or is the author just regurgitating what he's heard elsewhere?
"On already maxed-out servers, virtualisation will be a total waste of time" - this couldn't be further from the truth. Maxed out servers are more likely to experience component failure due to the consistently high workload on those components, and would therefore benefit extremely well from virtualisation since they will gain a high level of resiliency/portability/flexibility (assuming shared storage is used, which is a given with any even half-serious virtualisation project), than if they were left as a physical box. There are also other benefits for these servers, such as being able to dynamically provision extra resource to better cope with increasing workloads over time, without having to take the application offline etc.
"as much as customers like more cores – especially in virtualized environments, where they tend to pin one virtual machine on one core for the sake of simplicity"
I'm not sure how much hands-on virtualisation experience the author has, but I can honestly say that in my 5+ years working with enterprise grade VMware environments, I've hardly ever seen customers do what is suggested above. This goes against the whole concept of what virtualisation is about, and would dramatically reduce the number of VM's you could realistically fit on each host, not to mention seriously undermining the flexibility that you would otherwise get from the environment.
"It caused recurring weekday alarms for the iPhone to ring one hour late on Monday"
Wrong way round, the clocks went _back_, meaning an extra hour in bed, so the alarm would have gone off an hour too soon, which would mean a "rude awakening"... surely an alarm going off an hour _late_ would be a 'happy, relaxed, refreshed awakening"!?
Also, wasn't this reported a couple of weeks ago when our clocks went back [in the UK], and the same mistake was made then as well... sloppy ;)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
