Bladewatch.com

I got an email from Mike asking about his Compaq DL380 server, the disk has failed, they called an engineer out and he’s swapped it.  How does he know if the drive has rebuilt? I’ve included my response below:

Hi Mike,

Not a problem, your engineer should have checked it rebuilt before he left, but typically the drive should automatically rebuild in about half an hour or so, it depends on the size, file structure and how busy the server is. (There is also a setting on the controller to set this), the same principles apply to HP servers, though the agents are renamed and updated.

There are a few ways to tell if your drive is rebuilding:

• Event viewer on Windows – it will show drive failure, drive removed, drive inserted and drive rebuilding, once it’s rebuilt there will be an entry saying the drive status is now rebuilt.
• Compaq Array Utility – this is part of SmartStart drivers/applications, run this tool and it will show you a diagram of your drive status
• Compaq Web Agent – if you’ve installed the Compaq web agent, on the server, you run http://insertyourservername:2301, then log in, it will show you if the drive has rebuilt, by basically showing you the array controller status on the web page.

Basically using the compaq tools, everything should be green and there should be no errors on screen, you could also check the compaq integrated management log, though if memory servers, it doesn’t state drive rebuild, just failure.

Physically look at the front of the servers on the drive. If the drive is rebuilding, the drive should be flashing to illustrate it’s rebuilding the array or disk.

Hope that helps.

Regards

Martin

Share and Enjoy

I got an email from a small organization that shall remain nameless.  They’re working on a new storage solution which they’re quite excited about (and rightly so), they were asking once they’re ready how do you approach the different target markets, namely SME and hopefully enterprise customers, they were asking the following:

• How would I sell it?
• How would I put it into the enterprise market?
• How do you sell to SME?

Some good questions, I have to confess that I am not a sales person, but let me answer you this way.

How would I sell it – key issues

• What are the technical requirements
• What technologies or standards do I need for your product – iSCSI, 10GB Ethernet, fibre etc
• What technologies does the platform support – Mac, Windows, Linux or Unix
• What key benefits does it offer, both in terms of cost, delivery and support?

How would I sell it in the enterprise?

• How will it work in a structured/owned environment – ‘Storage’ owns the NetApp filers, the EMC, the SAN – how would it fit in
• What license and support costs are there and what is included?
• What benefits in operational costs, delivery and performance can be achieved
• Is it a cross platform, cross business tool or something that say only finance might need – who would buy it, the CIO? CEO?

To SME’s you’re looking at

• Reducing cost
• Ease of use and support – I don’t need a guy to look after it
• Compatible with other technologies

Consider a blog for the product/organization but key:

• What the product aims to achieve – and how you’re doing this
• What issues have you faced with the product
• What benefits have you recevied – expected and unexpected
• What customers have said and acheived with the product

Share and Enjoy

By Jake Sorofman

Oracle’s bid for Sun has officially sucked the air out of this week’s news cycles.
And for good reason—it has the potential to truly change the face of the IT landscape, pitting Larry of the Emerald Towers against the stalwart incumbents of the enterprise datacenter. This is a story we’ll all follow closely for some time.

In other news this week, rPath announced its plans to redefine application delivery and systems management. Certainly, I’d be disingenuous to suggest that I’m not at least slightly lamenting the timing of our news, which—while, important—may struggle to rise above the din of the Oracle/Sun announcement.

But I actually see an important coincidence in the timing of these events.

You see, rPath believes deeply that IT has lost its way and has become far too focused on infrastructure, rather than applications. As a company, rPath has set out to reorient the world to the notion of application-centric—rather than server-centric system management. In this model, operating system, middleware and tooling are subservient to the application, which becomes the unit of management.

Check out this post which is on the rPath site written by Jake, it’s and interesting read, and raises a point which I sometimes have to walk a fine line between as a ‘blade boy’. Ask any business/application buy, the person paying the bills, you’ll find something interesting:

“I don’t care what platform it is…. It’s got to be high availability…. The cheapest way please…”

I’ve paraphrased a conversation I was having only last week with a head of business unit, when asking him about blades/virtualization obviously it depends on your business, your comfort and exposure to the technology, but as an IT Professional we need to:

• Maintain delivery – avoiding unnecessary outage, protecting and stabilizing the core infrastructure – the network, the storage, the server.
• Maintain a stable platform – configure with excellence in mind, provide a secured, auditable platform ready for business.
• Aid application performance through effective best practice and configuration

The platform, be it a Compaq Proliant from the olden days, a blade or a desktop with 128mb ram, as long as it works, it’s stable and delivers, the rest is just noise.  Does IT energy efficiency matter? Does what platform you choose, x86/SPARC/Power etc matter? The operating system? To the end user the guy pressing submit no. But these kind of questions do underpin the ability of IT to deliver a ‘stable core’ for grid/hpc or application virtualization, but crucially END USER DELIVERY.

Share and Enjoy

Enterprise Storage Forum

Today’s surprising news that Oracle (NASDAQ: ORCL) will acquire Sun Microsystems (NASDAQ: JAVA) means that the data storage industry suddenly has a new — and formidable — player.

“Our customers have been asking us to step up to a broader role to reduce complexity, risk and cost by delivering a highly-optimized standards-based product stack,” Oracle president Charles Phillips wrote in a letter to partners and customers. “Oracle plans to deliver these benefits by offering a broad range of products, including servers and storage, with all the integrated pieces: hardware operating system, database, middleware and applications.”

The deal is much more likely to clear antitrust hurdles than Sun’s proposed merger with IBM (NYSE: IBM), which fell apart two weeks ago.

But Wedbush Morgan analyst Kaushik Roy doesn’t necessarily think that’s a good thing.

“IBM was a better strategic fit for Sun,” he said. “Now Oracle is getting into totally new markets in which they have no expertise or history. I am skeptical.

We’ll have to see how the deal goes forward, until any specific announcements are made, that innovation of the product and platform continues is key, as is the ability to meet customers expectations. We need not worry about hot the organizations fit organizationally, at the moment the key things are:

1. Understanding how the new organization will operate going forward
2. What products/platforms continue to be innovated or are retired
3. How open source projects fit within the new organization
4. Focus on the business as usual – maintaining revenue, maintaining customer focus and avoiding uncertainty. This is key.

Share and Enjoy

Real wire

London, 21 April 2009 – The organisers of the 5th Data Centres Europe Conference and Awards have said that despite the downturn, attendance figures have been sustained from last year with more than 300 executives – some 25% of which are at CXO level – converging on the two-day event which begins tomorrow in London. (www.datacentres.com/dce)

5th Data Centre Europe is the leading regional forum for data centre, managed services, outsourcing and cloud computing businesses. The conference brings together Data centre operators, service providers, telcos, network infrastructure providers, managed services specialists, systems integrators, enterprise and government IT directors, property and investment firms, law firms, professional intermediaries, consulting firms, regulatory bodies, and supplier organizations to inform and update, provide critical insight and offer top level networking opportunities.

Data center space and power continues to be an issue, whether you’re an organization that’s reaching it’s operational limit, or if you’ve acquired another business and have too much space. How you evaluate your data center, how it affects your business your revenue remains a key part of your IT infrastructure strategy. It will be interesting to read the different speeches as well as see what the panel discussed and talked about, I wish I’d been able to attend.

Share and Enjoy

Computerworld

April 14, 2009 (Computerworld) EMC Corp. today unveiled a high-end Symmetrix storage array that is based on Intel Corp.’s x86 quad-core processors and integrates VMware’s APIs to automate the provisioning of storage for virtual machines across multiple tiers of disk drives, from solid-state to SATA.

The new V-Max array, with its Symmetrix Virtual Matrix Architecture, scales to 2 petabytes of usable capacity within a single chassis and offers three times the performance of the current Symmetrix DMX-4 array.

EMC said that its use of industry-standard processors and VMware’s hypervisor enables the system to scale up to hundreds of thousands of terabytes of storage and tens of millions of I/O operations per second supporting hundreds of thousands of VMware and other virtual machines in a single, pooled storage infrastructure.

Anything the vendors can do to make storage more affordable and accessible has to be a good thing, that EMC are also improving the performance and scalability of their storage brings more opportunities to the virtualization platform.

Share and Enjoy

HP.com

HP today unveiled new solutions within its Adaptive Infrastructure (AI) portfolio that fundamentally change the way technology is used to deliver business services, while also reducing overall infrastructure costs and data center complexity.⁽¹⁾

HP BladeSystem Matrix (Matrix) is the industry’s first converged software, server, storage and networking platform that automates service delivery for the data center. HP Matrix Orchestration Environment provides a unified management interface to rapidly design, deploy and optimize the application infrastructure.

Together, these offerings create an integrated pool of resources that operate in both physical and virtual environments, creating the first truly business-ready infrastructure.

This does sound cool, an all in one solution for managing your infrastructure including the storage, network and server could bring real savings in the time to deploy new systems, and enable delivery through on-demand provisioning of infrastructure using user pre-defined templates. For that Citrix farm, I could deploy another server, network connection and storage as the application brought on new users to meet existing levels of availability or service. Simply handing over elements of the provisioning to ‘the business’ can transform the perception of the IT teams, and allow more focus on upgrading the legacy platform, pro-active monitoring and evolution of the infrastructure through capactiy planning. Interestingly, HP BladeSystem Matrix supports virtual and physical servers on multiple operating systems, it will be interesting to see what people think about this new announcement.

Share and Enjoy

Market Wire

SANTA CLARA, CA–(Marketwire – April 20, 2009) – BLADE Network Technologies, Inc. (BLADE), the trusted leader in data center networking, will present viewpoints about 10 Gigabit Ethernet, cloud computing and network virtualization for the dynamic data center at BladeSystems Insight at the Mandalay Bay Resort in Las Vegas, Nevada on April 20-21.

Dan Tuchler, Vice President of Strategy and Product Management, will be a speaker on the Networking and Infrastructure Panel: “Is now the time to move to 10GbE?” moderated by James Staten of Forrester Research, along with panelists from Broadcom, Cisco, Intel and Juniper Networks.

Tuchler will also be a participant on the Virtualization Recipes for Success Panel: “Real World Stories from Today and What’s Coming Tomorrow,” moderated by Barb Goldworm of Focus Consulting, along with panelists from Cisco, Fujitsu, INX and VMware.

BLADE executives will also present case studies and examine the networking requirements for cloud computing in two customer presentations.

BLADE will make available to BladeSystems Insight attendees a Networking Brief developed by Enterprise Strategy Group (ESG) entitled “Enabling VM Mobility via Intelligent, Automated, Virtual Machine Aware Networking Solutions.” The brief can also be accessed at: http://www.bladenetwork.net/Analyst-Reports.html.

The brief does sound interesting, I wish I’d been able to attend the conference this year. I can certainly see the demand for 10GB ethernet as part of a solution for virtualization or hpc and hopefully as the cost of the adaptors and the switches falls, the adoption and innovation can continue to develop to bring new opportunities and possibilities with the infrastructure.

Share and Enjoy

Super Micro

BEIJING, China, April 8, 2009 – Super Micro Computer, Inc. (NASDAQ: SMCI), a leader in application-optimized, high-performance server solutions, is showcasing a 2U Twin² SuperServer with four hot-pluggable computing nodes in a 2U chassis and a multi-GPU SuperWorkstation today at IDF Beijing 2009, Beijing International Convention Center (BICC), booth GE014. These new systems are optimized to fully leverage the benefits of the Intel® Xeon® Processor 5500 (Nehalem) Series to deliver the best performance-per-watt and performance-per-dollar in the industry. Compared to previous generation Xeon platforms, Nehalem features more than double the CPU, triple the memory and triple the I/O bandwidth to deliver the largest increase in performance in the history of Intel’s Xeon product line.

I was doing some research comparing Super Micro servers with some other vendors and noticed this article talking about their 2u solution, which involves four hot-pluggable compute nodes in a 2u format enclosure, it does look cool. I’m off to read the energy requirements and specifications.

Share and Enjoy

Emulex

HP Technology@Work – BERLIN, April 20, 2009 – Emulex Corporation (NYSE:ELX) today announced that its LightPulse® 8Gb/s Fibre Channel Host Bus Adapters (HBAs) are optimized for use with the new HP BladeSystem Matrix, an out of the box, pre-integrated platform that creates pools of server, storage, power and network capacity that can be divided and allocated across up to 1000 physical or virtual servers. The Emulex LPe1205-HP 8Gb/s Fibre Channel HBA (mezzanine card)  is also available for use with HP BladeSystem c-Class and server blades, including HP ProLiant G6 and HP Integrity blades with support for HP-UX 11i, Linux, VMware, Windows Server 2003 and Windows Server 2008 operating systems.

Emulex LightPulse 8Gb/s Fibre Channel HBAs provide end user customers with the increased performance, I/O bandwidth and the scalability necessary to support next generation data centers. Emulex 8Gb/s storage area network (SAN) connectivity solutions also support the new HP Virtual Connect 8Gb Fibre Channel module and the HP StorageWorks EVA4400 Fibre Channel array.

Providing more choice in adaptors for the end user community has to be a good thing, particularly in terms of performance for Virtual Connect. The adaptor brings more bandwidth to support additional virtual machines, and is backward compatible with the previous adaptors, using a common driver model.

Share and Enjoy