Bladewatch.com

HP.com

HP expands upon the industry’s first mission-critical Converged Infrastructure with enhancements to the Serviceguard Solutions portfolio for Integrity and HP-UX operating environment.

Enhanced HP Serviceguard Solutions 11.20 availability software, combined with the recent release of HP-UX 11i v3 Update 7, enables clients to:

—    Reduce manual configuration efforts by approximately 93 percent⁽¹⁾ with the new HP Serviceguard Extension for Oracle e-Business Suite. It provides automated configuration services for more than 200 tightly integrated applications, including financial services, customer relationship management applications and interrelated components such as storage and networking.

—    Achieve improved database manageability with the new HP Serviceguard Toolkit for Oracle Data Guard. This solution is intended to provide robust high availability for Data Guard with zero custom coding given the need of Oracle database administrators for uninterrupted access to this critical application-management tool.

—    Simplify clustering with new graphical wizards in Serviceguard Manager. These graphical wizards can enable the creation of a complete cluster, including storage and networking, in about a minute. Additionally, the new Cluster Verification for HP Serviceguard provides confirmation that clusters are configured properly.

—    Restart multi-instance applications such as Oracle Real Application Clusters (RAC) and SAP up to 90 percent⁽²⁾ faster with enhanced Serviceguard performance. In particular, Serviceguard has been designed to load cluster packages simultaneously on all nodes instead of sequentially for rapid start up of the entire Oracle RAC stack, helping to ensuring critical applications are back up and running as quickly as possible.

HP continue to make progress in the concept of converged infrastructure, a topic that I know has many CIOs talking, more from a deployment and make it work principle than anything else. Anything the vendors can do to aid with the interactivity of the different layers of the infrastructure, to make reporting, management and application or infrastructure deployments easier and faster has to be a good thing for competition within the industry and IT as an enhancement tool rather than something which ‘needs fixed’.

Share and Enjoy

Check out this whitepaper, it’s an interesting read talking about the top 20 VMware performance metrics you should know about. Anything vendors can do to aid users with their platforms, whether it’s in terms of best practice or understanding features has to be a good thing. Do check it out.

Share and Enjoy

MarketWatch

ROUND ROCK, Texas, Sep 16, 2010 (BUSINESS WIRE) — Dell announced today new PowerEdge(TM) servers, PowerVault(TM) storage, and PowerConnect(TM) networking solutions that help growing businesses update their technology infrastructures with easy integration, minimal downtime, and the ability to scale — all while driving business performance under constrained budgets.

“Because of their limitation in IT resources, midsized businesses have always favored IT solutions that are: easy to install, configure and manage; are simple to use and require minimal customization; don’t increase demand on already limited IT resources; are easy to integrate with existing systems; and have straightforward and easy pricing,” said Jim Browning(1) of Gartner.

It’s always great to see product and platform innovation, anything Dell can do to aid customers in translating their business needs into reality has to be a good thing for competition and product innovation as well as the end user community. The new PowerEdge servers sound particularly interesting as do the storage solutions, I’m off to read more about them on Dell.com

Share and Enjoy

MarketWatch

ROUND ROCK, Texas, Sep 16, 2010 (BUSINESS WIRE) — –Over the course of five years, Dell blades can help customers reduce energy costs by up to $60,000 when compared with similar solutions(2)

–Energy savings can reduce environmental impacts for customers and the planet

–This is consistent with Dell’s ongoing strategy to eliminate data center inefficiencies through open, capable and affordable IT solutions

Customers in the Virtual Era require technology solutions that provide the right balance of performance, manageability and energy efficiency. Dell today released the findings of a study that demonstrates the company’s PowerEdge M610 blade servers with M1000e enclosure delivers better performance per watt and consumes less energy overall than HP BL460C blade servers with C7000 enclosure and IBM HS22 blade servers with BladeCenter H enclosure(1). Dell’s ultra-efficient blades can consume up to 20 percent less power per blade server than identically-configured competitive offerings(1) and help provide an ideal platform for environmentally conscious IT managers looking to reduce operating costs and the environmental impact of their data centers.

I was doing some reading about energy efficiency in the blade space and came across this article comparing the efficiencies and performance of Dell/IBM and HP servers, it’s an interesting read, as with any of these comparisons, the proof is how they perform in your configuration, in your space and your world, key drivers are building the right configuration in terms of processor, memory and power supplies, as well as looking at the air flow and utilization of the platform.

Key drivers for me in the blade space in terms of importance are:

• Manageability – utilities around systems management, configuration, deployment and support
• Energy efficiency – the concept of performance per watt, or midrange processors, maximum performance at the least marginal cost
• Performance and compatibility with external platforms – ESX, SAN storage, 10GB Ethernet, technologies that I might not need now, but might be considering going forward

Share and Enjoy

I was having a chat with Angus we’ll call him asking him how he was dealing with a new role working for a medium business; it has four sites across the UK in the financial sector with external clients, a set of complex infrastructure and a server estate in a rapidly changing state. The conversation was over Diet Coke and coffee and was unscripted.

Tell me about your business

We’re an expanding business, we’ve got just over two hundred users and the business is planning to purchase additional businesses or units when the chances present themselves. They’re looking to expand, as a result our server estate is kind of all over the place, from the original business we have a degree of standards, they might only buy one type of server and model, from the businesses we have bought, and we have a range of vendor’s equipment and configurations.  We have a range of in-house developed and bought in applications, and we have systems that we use and that external customers use under contract, when they go down we loose reputation and sometimes have to pay penalties, so that can be a problem.

Tell me about the role

My role has been brought in to improve service in a mini service delivery/server manager way. There is no windows/unix server team, the guys look after all the servers we have and help out the desktop guys when they get into difficulty, however there are a number of problems. Fundamentally the server team had effectively started responding to emails and abstracting themselves from the helpdesk tool, this meant a build up of calls which kept growing to the point that when they opened the helpdesk tool it had hundreds of calls and requests which the engineers then felt was unmanageable, thus they continued down the email trail. Management kept looking at unresolved and unresponded calls reports questioning what was going on. The reliance on email caused problems for the end user as they might log calls and have no one respond to them, the teams that used calls then had to start switching to email taking the work log communication, the status updates and information out of the call meaning we couldn’t track issues and identify problems centrally – it always relied on asking Mike or Janet.
Additionally with the mergers and integration activities we had servers all over the place from a support and deployment process, if one business unit asked for a new server it had to be built in a specific way and meet specific criteria in certain domains within their active directory, backed up and managed by their set of tools, making everything overly complex.

So what has been your focus?

Firstly an inventory of the systems and applications we have so that we understand where we are and what we have to manage, where our operational risks are, examples being Joyce who stores her accounts on a Compaq Proliant 1600r below her desk (we’ve still to work out why), or the shared file server which is a DL380 G4 with several array shelves attached to it, in large volumes making management sub optimal and offering a central point of failure for all business units. Boys, is there a bcp equivalent? Well we have it backed up was the response, but no one had questioned the time to recover nor reliability of the backup.

With the inventory in mind, we then moved towards standards, with a focus on the simple things, I’ll give you an example, one business unit only had thirteen servers so they were very impressive, everything was configured just right, firmware/drivers were up to date and everyone knew everything about them, the other business unit had 80 servers, so their estate was a bit more complex, the support information was in document upon document, and the servers were all at different levels from operating system to firmware/drivers and middleware.

We therefore developed a series of rules:

• Windows NT was no longer in support – all servers had to be rebuilt Windows 2003 within three months (we’ve just completed that using VMware and physical servers where necessary)
• The versions of Solaris supported were reduced to current and legacy, anything legacy had to be version 9 to keep things simple from a support viewpoint
• All platforms to have a development equivalent to reduce outages and improve testing
• All platforms to have a basic driver/firmware matrix which we set to keep up to date, this includes the middleware applications like SQL etc
• Reduce the range of hardware platforms from 19 to four, upgrading as we go, you either have small, medium or large, DL100/DL300 or DL500
• Backup inventory taken – what was backed up against what was supposed to be backed up
• Servers to have lights out functionality as standard for remote support and to allow us to consider data center consolidation
• Reduce the range of operating systems from five to three, Solaris, Red Hat Linux and Windows.

Additionally we had two projects started simultaneously, with separate technical teams for each.

• BAU2 – Effectively I took four of the nine guys and said fix production, stabilize it and make it work.
  • Stabilize what we have, the server guys had this viewpoint (as did some of the management team) that we would address that in version 8.21, when we moved to Windows 2003, or when we fully integrated that satellite site off Lotus on to Outlook. The problem with this was that we continued to spend most of our effort on keeping things online, that the Windows NT servers were running out of disk space, caused calls, emails alerts and out of our escalation, not to mention outage for the end users, so we established the set of rules I’ve mentioned to maintain production.
  • With a stable legacy platform in terms of support or environment (Windows NT) or an integrated business lines, we could then focus on integrating, standardizing and consolidating, with the applications working reliably (to an acceptable level), we could then allow the app teams the time they needed to migrate onto new physical or virtual servers, even start migrating users on to this Outlook thing, ready for Gen3 where we buy in our email from google, from an Exchange cloud etc.

GEN2, phase 1 – The other engineers were then dedicated to service improvement through renewal, refresh and consolidation or integration:

• Integrate and consolidate the back office IT functions, so there is one print environment and configuration, one email platform, one file server per business line, until we move towards a centralized active directory (what we refer to as GEN2 of the integration and IT platform)
• Virtualize old kit until it can be upgraded, this allowed us to more easily adapt to business need and save on data center space and hardware costs, and the support costs alone for some of the legacy x 86 servers could pay for a new server.
• Network and back office Windows integration and consolidation were kicked off, a move towards a common way of working, away from host files if they were used to dns, a standard way of creating accounts, everything in active directory, processes to decommission users, printers and computers, so that going forward we kept everything tidy.
• With the desktop team move towards a unified build, a unified standard for desktop so that things like Citrix and the layered applications converge, moving along simple concepts, once your pc is three years old it gets replaced with a Gen2 build (Office 2003, Windows 7 in our new AD).
• Simple things like standardize printers, in our North of England office someone there thought Xerox was the printer of choice, Glasgow liked HP and London seemed to buy whatever was on offer that month as the printers were bought by the department. Simply reducing the range of printers to two, either colour or black and white reduced the complexity of the environment and the trauma of managing drivers. Purchasing guy could then debate server and software costs with our service providers rather than the toner price for a LaserJet II for the office in Glasgow for Gerald who didn’t quite want to buy a new printer.

What about architecture and evolution?

This was another problem when I started we had a lot of dreaming and vision, a lot of what if statements all mixed in with the by phase 3 that will be fixed. We had far too much of the aiming for the universe but neither delivering it, nor stabilizing or improving what we had. I’ll give you an example, one of the guys from the architects had allocated an engineer and network guy to go off and test 10GB Ethernet for our environment in London, on the basis that if we implemented grid computing (which we’ve been testing) or scaled up our virtual environments 10GB Ethernet was what we needed. The issue was that they weren’t wrong, we need to be looking outside the box, thinking of how we can do things better and what technologies we should be looking at, but right now, that’s not a priority. Our priority was the seven year old server which died taking us offline for three hours resulting in penalties in the five figures, that’s real money, real end user and business impact. Simply stabilizing that platform or moving the application to a new server might result in real change and real opportunities to improve revenues and profitability.

What else did you work on?

• Everything in the helpdesk tool space, when the team gets emailed, the team know to forward it to our helpdesk team to be logged
• Every engineer goes through the queue and closes down redundant calls, re-logging outstanding issues so that we know where to begin with three engineers ready just to sit and go through new/existing incidents
• The Service management and help desk manager to go through all desktop calls and the remaining queues identify old calls and issues; highlight them to the management team so that they get resolved appropriately.

How is it going?

• We’ve reduced production outages and incidents by 30% which is impressive and thanks to a lot of hard work from the BAU2 team, giving four guys the chance to own everything, make the decisions has freed them in some respects, they feel they own the infrastructure and are accountable for it, there are clear lines of engagement for issues, everyone knows the BAU teams number and who to ask for.
• With the integration and consolidation we are also making significant savings simply on the standard small items and our hardware support contract. We turned around the our support provider and agreed anything older than three years was out of support, that reduced the costs significantly and allowed us to transfer that saving to buy new servers, it also allowed us to breach the idea of what actually needed support and what level of support we needed.

Standardizing the nuts and bolts on the desktop and the server side has meant more time to look at how we are working and less lead time for equipment, it’s now one of three types of printer, one type of desktop, one of three pre-agreed server types so quotations, purchase orders are a lot quicker and easier. Working on things like the server build workflow to remove the redundant tasks and prioritize the ones that matter, the inventory being updated at the right time, setting task dependencies and setting automated notification to remove delays made a big difference. Consolidating the back infrastructure did cause some initial issues as we worked through legacy configurations and platforms, but we upgraded service as we moved, once the message got around, business units were then motivated to join Gen2 of the infrastructure as we called. One of the most effective ways of achieving that was if you got a Gen2 pc, it came with a user drive of 1GB, it came with windows 7 and the new office, not to mention web based email and departmental drives that had fewer restrictions on space. Simply migrating shares on to newer servers allowed for more space, and by getting the users to want to move on to the new infrastructure, meant them asking when are my apps moving and why haven’t they moved, again creating an incentive to move on to the newer standardized infrastructure again, reducing the legacy overhead and making teams adopt new standard to protect them and the infrastructure going forward.

Share and Enjoy

http://www.thebdevent.com/paloalto.shtml

PALO ALTO, CA – September 20, 2010, TheBDEvent, a networking event created to help technology executives develop the “business behind their business,” today announced details for its upcoming January 2011 conference, to be held January 25th – 27th at the Sheraton Palo Alto (CA). Registration is now open and is limited to qualified senior executives in the storage, networking, security, virtualization and cloud computing industries.

TheBDevent offers educational, networking and business development opportunities without the sales and marketing distractions of typical trade shows. Ed Walsh, CEO of Storwize (recently acquired by IBM) and formerly the CEO of Avamar (acquired by EMC) and CEO of Virtual Iron (acquired by Oracle) will kick off the event with a presentation on the tactics, do’s and don’ts of successful top executives.

“I have participated in past BDevents and have found that having a laser-focused, dedicated event specifically for executive networking is a great value to me,” said Walsh. “I look forward to sharing some management and business development strategies learned over the past few years.”

Other Agenda highlights include:

• Keynote presentations by industry luminaries
• Matchmaking Application – the BDevent has created a unique, custom application for quickly and accurately matching your needs to other attendees and the types of products and services they offer. This graphical application has been specifically designed for busy executives who want results fast. There are no other technology trade shows or events that can put executives together in such a meaningful and fast way. Find who you want, when you want – right up to and throughout the event itself.
• Venture Capital Workshops, Presentations and Speed Dating (five minutes each with notable Venture Capital firms such as Highland Capital Partners, Valhalla Partners, Intel Capital, Jafco Ventures and others)
• Analyst and Press Briefings Panel – present your company and products to panel of analysts and press, gaining valuable feedback and insight from prominent industry veterans. Panelists include representatives from Storage-Switzerland, Storage Strategies NOW, InfoStor, DCIG, Storage Magazine and SearchStorage
• Marketing and BD Forums and Roundtables

This sounds like an interesting event, it’s always great to see new developments in storage, whether its in terms of de-duplication, process or providing more dynamic and energy efficient storage platforms for the different tiers of data within the enterprise. I wonder what topics the keynotes will cover, if there is anything about data storage issues in the enterprise as well as concepts like virtualization, I’ll need to read up more.

Share and Enjoy

HP.com

Clients can deploy applications in hours

SAN FRANCISCO, Sept. 20, 2010 – HP today announced integrated private cloud solutions customized for Oracle applications that significantly shorten and simplify the deployment process.

HP Private Cloud Solutions for Oracle Applications are specifically engineered to support complex, multi-tiered environments, so clients can deploy Oracle applications in hours. The HP hardware and software offerings ship together from HP Factory Express pre-optimized for PeopleSoft Enterprise, Oracle E-Business Suite and Oracle Fusion Middleware. Based on HP Converged Infrastructure, the turnkey solutions are easy to buy and install, eliminating guesswork, weeks of configuration time, and minimizing architectural design effort.

The new HP Private Cloud Solutions for Oracle Applications are based on HP BladeSystem Matrix and enhanced with HP Cloud Service Automation software and HP StorageWorks data services. The solutions offer one-touch application provisioning as well as a self-service portal for delivering Oracle application services on-demand.

With a common management environment for the integrated server, storage and networking infrastructure, clients can significantly reduce maintenance tasks and costs. They also provide automated deployment, monitoring and compliance of Oracle applications.

Through HP Cloud Maps customized for Oracle applications and imported directly into HP BladeSystem Matrix, clients can rapidly populate data and build a catalog of application services for the business.

HP Cloud Consulting Services provide the expertise needed for clients to transform their existing delivery approaches into more efficient shared-services models, including discovery, design and implementation.

“To ensure the best service delivery across an organization, clients need applications and infrastructure to be tightly integrated and IT systems to be tuned to the unique demands of each application environment,” said Mike Crowsen, vice president, Oracle Alliance, HP. “No other vendor has integrated infrastructure for Oracle applications to the extent HP has with its Private Cloud Solutions for Oracle Applications, enabling clients to achieve levels of deployment speed, simplicity and flexibility never before possible.”

“Clients’ application service delivery should be streamlined and simplified, and services delivered more efficiently and securely,” said Judson Althoff, senior vice president, Worldwide Alliances & Channels and Embedded Sales, Oracle. “HP Converged Infrastructure – in particular the new HP Private Cloud Solutions for Oracle Applications – delivers an open, common architecture, managed in a common way that can be optimized for the needs of any application, in any environment.”

We are seeing more developments in the cloud space, whether it’s through managed services, or customers deploying the right set of infrastructure and middleware applications to make application deployment and maintenance that bit easier and more efficient. Anything HP can do to assist with Oracle applications and platforms has to be a good thing for the end user community and HP alike, I’m off to check out more.

Share and Enjoy

I got an email from a local small business manager asking about how to do a driver and firmware upgrade, the email text is below, I thought I’d post it, the process is the same for the other Proliant servers, as always do ensure you read the release notes for firmware and drivers, also consider having a back up, though just in case – this is more as a liability statement than anything else. The links are in this case for a DL380 G2 as this was John’s server, although the concepts are the same.

Martin,
My IT guy is telling me we need to upgrade the drivers and firmware for our Windows 2000 file server, I was wondering if this is something I can do rather than call him in to do it to save money?

Is it complicated and do I need to do specific steps he made it seem quite difficult. It’s a DL380 G2?

Regards

John

The answer is that it’s not overly complicated, there are a few steps or considerations, the only thing to note is if you have pc anywhere/remote control option installed on the server as these can conflict with upgrading the driver pack, but I’ve written that here. This page contains links to documents showing the firmware upgrade process.

So here is the mini plan I put together for John:

1. Reboot the server (just to make sure that it is fine before you make any changes)
2. Download system firmware – this is available here – When it asks to save, I normally say yes and call it the server model, so DL380.exe
3. Download the array firmware – this is available here – Again, I normally say yes to the save and then call it smart.exe
4. Now we need to download the Proliant Support Pack – this is available here – this contains all the drivers and utilities – it’s here
5. Now copy all the files to the server, c:\temp for example
   1. Now log on to the server and browse to c:\temp and do the following:
      1. Run dl380.exe (the system firmware you just downloaded), the utility will tell you the current version and ask if you wish to upgrade, say yes and proceed. Now it will ask for a reboot, I’m going to select no to reduce the number of reboots – you don’t have to.
      2. Run the smart.exe (the firmware you just downloaded), again you get the standard message, current and upgrading to version, say yes and proceed, when asked to reboot, reboot the server.
      3. It will then reboot, the date of the bios will change and if you look closely the array controller will load with a new version or date as well
      4. Windows will now load as normal. However it will possibly prompt to say that new hardware has been detected and do I want to reboot. I say no at this point, let everything continue to load up as normal, and then reboot.
6. Now we run the driver pack (HP Proliant Support Pack). Run the exe that you have downloaded and extract the files to a folder for example c:\temp\drivers, now you can either run the setup which guides you through the process and then asks you to reboot, or run the cmd file which will install everything, the cmd will then close and then you reboot.
7. From start to finish, it wont take more than about forty minutes including reboots.

Share and Enjoy

I had an issue at a client site where they had run the Proliant Support Pack (the driver pack) on their Compaq Proliant DL380 and when the server came back online, it was stuck in vga mode, 16 colours and would not allow the user to change the resolution. We did some more investigation and found that Remote Control Option, VNC and Symantec PC Anywhere were installed on the box, with one of them causing an issue – it’s on the HP Support forums, but this is the quick way to fix it.

• Stop and disable the remote management services (VNC/PC Anywhere/Remote Control Option) and reboot.
• The server will come up online in vga mode which is fine.
• Right click my computer and do manage
• Device manager
• Select the VGA driver and select the previous driver then reboot.
• Log in to the server and set the resolution you want 800×600 at 256 colours for example
• Set services you disabled to automatic and reboot.

Share and Enjoy

I had been speaking with a local SMB and a head of IT for a bank asking them about their servers, their current top five issues and it was interesting to hear that although they operated in different spaces, different markets and budgets, their needs were beginning to align from an end user standpoint, I’ve removed any references to the companies that they work for to avoid any debate, but the results remain the same.

SMB your top five issues

1. Servers are coming out of support – do we buy new ones or run the risk, do we look at a support contract, but then we don’t like the cost of doing that
2. Everything takes too long – why does replacing my web server become a three month project? Don’t we get a Gerald, with a linux dvd, a server and a network cable?
3. Complexity of the infrastructure – we’re getting more layers within the IT space as we look at more solutions, more interdependent things which can be troublesome to troubleshoot when an issue arises – the SAN went on its holidays for a few hours last month and we had to call in a guy to reset it
4. Loyalty and relationships – we want a good deal but we also want a degree of flexibility and someone that we can call at 3am to ask advice – we want buy in and a customer relationship which might not comprise of everything from one vendor
5. Support getting quite difficult and expensive – our DL380 G2 system board failed, our service provider quotes us £1000 for a system board, £350 to fit it plus call out fees, we ended up getting an engineer to buy one from ebay for £95 and fix it himself.  We would have paid £300 to the service provider if they’d had one, we needed the bits to fix it for the next few months whilst we migrate/fix it

Mr Enterprise your top five issues

1. The need for support costs, yet the lack of interest in paying for them due to the cost, being able to know where the line is between what needs to be on the contract and what actually is, coupled with being able to take money from the support contract to upgrade old hardware
2. Complex infrastructure, as we start with different application tiers, more middleware and infrastructure layers like clustering, SAN and NAS, scheduling and monitoring tools, the time taken to fix a problem can seem like longer and involve more teams
3. The pressure to absolutely look at the best deal to save money whilst not appreciating the internal dynamics and associated direct/indirect costs of doing so, getting let down by a vendor over a trivial issue changing relationships from on the radar to high maintenance
4. Challenges for a more diverse support model, business users wanting more involvement and control of the infrastructure, taking on more of the nuts and bolts infrastructure support, and the complexities in doing so, we aren’t necessarily geared up for hardware support and for commoditized platforms, saying rip and replace is fine in the event of failure, but only if the business sponsor has signed off on the risk of doing so. The key example, our Citrix x86 server which died four months outside the warranty, we needed a new backplane and some memory, but instead had to buy a new box and migrate resulting in weeks of outage, rather than four man hours and parts.
5. Pressure around the data center space, in being more adaptable and responsive to change – why we aren’t looking at cloud without necessarily knowing what we would use cloud for, about in essence IT not having the answers for when someone wakes up and says right, we’re moving the email to cloud and we’re buying in grid compute resource.

Share and Enjoy