Or do they? And what sort of premium does the TCO for these platforms attract? IBM have been selling expensive high availability, low failure rate boxes for years, so this shouldn't present a problem!
Stating the obvious, but cost benefits to me Mr. Customer should come in the form of reduced maintenance costs (you've made me a box that heals itself, so I don't need so much of your technicians coming to fix it, and don't need them so urgently) and reduced front line support costs through call avoidance - (if it broke and fixed itself, why would my clients be calling me?)
Self healing servers hey. Firstly - just to clarify the 25% stake of their R&D budget they are wagering on this technology. The artcile mentions 25% of their server R&D budget will be expended on this technology. Not sure how much of the total R&D budget the server budget represents.
There has been a common trend towards reducing service requirements for applications and infrastructure for a few years now.
In the desktop market, Motive corporation (www.motive.com) have begun packaging their suite of products on Compaq, Dell and Gateway platforms. Motive's suite of products promotes a 'self-help and self-heal' service model. An agent application running on the client workstation can identify a fault with the operating system or any known application and offer resolutions - often in the form of restoring an application or operating system compoennt to a previous configuration state. Both in non-connected and connected modes. Quite neat!. Business logic can also be applied to try and understand what level of failure the operating system experienced and route directly to a third party maintainer where it has been deduced a hardware component is at fault or close to failing - A cheaper support model - the client is assisted with resolving their problem - one less call to my help desk!
So how will these things actually heal themselves? Would it be an extension on what most server boxes can currently do today - An operating system that has knowledge of states of various components can easily report their failure or impending failure.
What is IBM going to provide us? a self-healing model whereby the operating system can identify suspect hardware components and then isolate them as an unavailable processing resource? Given all these component resources should be hot-plugable, the IBM tech can come out and replace when convenient?
NSM tools can automate and avoid some operating environment issues. This should be possible on any platform, regardless of the inherent hardware healing capabilities. They could remember configuration states and know where to get a recent configuration to remedy a situation.
The theme among all of these replies seems to be the same - handsandfeet work might be reduced, but planning for and effective use of high availability platforms will still require skilled resources. Computers don't and won't engineer themselves!
When these servers come to fruition, I will be interested to learn more. Even IBM have said the technology is not new, but they are simply extending on what they already know.
One thing is for sure the 'Midframe' market (as Sun has coined it) will certainly be a competitive one. Sun, IBM (and I believe HP's Superdome offers or will offer similar functionality?) all want a piece of your Miracle-Enabling-Server-Budget
But I'll give you any money, that had you offered to pay the repair centre some extraordinary maintenance uplift on top of the warranty, that guaranteed you'd have a kitchen stove fixed within 2 hours, they would of taken your money!
Even if it did end up costing you more than 2 stoves over the general life of a kitchen stove.
This is one of the benefits of this transmission fabric not pointed out in the RFC.
The process of natural evolution will ensure the next release of the transport will include organic modifications correcting design flaws.
Unfortunately, this approach lends itself to comparison to Microsoft's product upgrade, new release and patching process. Especially considering the common user has no avenue through which to suggest enhancments or design flaws.
Perhaps the Chinese Government is just a bit perplexed there may be a people's revolt away from the commonly consumed Chinese Tea to a Coffee variant?
If worst came to worst, China would of course take the approach of meticulously filtering content as required. Here in Australia, we would probably just go, stuff it, there is too much stuff to filter, lets presume most of it is bad, and we'll pull the plug.
Slashdot Mirror
Or do they? And what sort of premium does the TCO for these platforms attract? IBM have been selling expensive high availability, low failure rate boxes for years, so this shouldn't present a problem!
Stating the obvious, but cost benefits to me Mr. Customer should come in the form of reduced maintenance costs (you've made me a box that heals itself, so I don't need so much of your technicians coming to fix it, and don't need them so urgently) and reduced front line support costs through call avoidance - (if it broke and fixed itself, why would my clients be calling me?)
Self healing servers hey. Firstly - just to clarify the 25% stake of their R&D budget they are wagering on this technology. The artcile mentions 25% of their server R&D budget will be expended on this technology. Not sure how much of the total R&D budget the server budget represents.
There has been a common trend towards reducing service requirements for applications and infrastructure for a few years now.
In the desktop market, Motive corporation (www.motive.com) have begun packaging their suite of products on Compaq, Dell and Gateway platforms. Motive's suite of products promotes a 'self-help and self-heal' service model. An agent application running on the client workstation can identify a fault with the operating system or any known application and offer resolutions - often in the form of restoring an application or operating system compoennt to a previous configuration state. Both in non-connected and connected modes. Quite neat!. Business logic can also be applied to try and understand what level of failure the operating system experienced and route directly to a third party maintainer where it has been deduced a hardware component is at fault or close to failing - A cheaper support model - the client is assisted with resolving their problem - one less call to my help desk!
So how will these things actually heal themselves? Would it be an extension on what most server boxes can currently do today - An operating system that has knowledge of states of various components can easily report their failure or impending failure.
What is IBM going to provide us? a self-healing model whereby the operating system can identify suspect hardware components and then isolate them as an unavailable processing resource? Given all these component resources should be hot-plugable, the IBM tech can come out and replace when convenient?
NSM tools can automate and avoid some operating environment issues. This should be possible on any platform, regardless of the inherent hardware healing capabilities. They could remember configuration states and know where to get a recent configuration to remedy a situation.
The theme among all of these replies seems to be the same - handsandfeet work might be reduced, but planning for and effective use of high availability platforms will still require skilled resources. Computers don't and won't engineer themselves!
When these servers come to fruition, I will be interested to learn more. Even IBM have said the technology is not new, but they are simply extending on what they already know.
One thing is for sure the 'Midframe' market (as Sun has coined it) will certainly be a competitive one. Sun, IBM (and I believe HP's Superdome offers or will offer similar functionality?) all want a piece of your Miracle-Enabling-Server-Budget
But I'll give you any money, that had you offered to pay the repair centre some extraordinary maintenance uplift on top of the warranty, that guaranteed you'd have a kitchen stove fixed within 2 hours, they would of taken your money!
Even if it did end up costing you more than 2 stoves over the general life of a kitchen stove.
This is one of the benefits of this transmission fabric not pointed out in the RFC. The process of natural evolution will ensure the next release of the transport will include organic modifications correcting design flaws. Unfortunately, this approach lends itself to comparison to Microsoft's product upgrade, new release and patching process. Especially considering the common user has no avenue through which to suggest enhancments or design flaws.
Perhaps the Chinese Government is just a bit perplexed there may be a people's revolt away from the commonly consumed Chinese Tea to a Coffee variant? If worst came to worst, China would of course take the approach of meticulously filtering content as required. Here in Australia, we would probably just go, stuff it, there is too much stuff to filter, lets presume most of it is bad, and we'll pull the plug.