Bladewatch.com

http://h20000.www2.hp.com/bc/docs/support/SupportManual/c00264524/c00264524.pdf

Check out this document which contains information about how to upgrade system firmware on HP servers using the SmartStart media (offline) and the online firmware upgrade process where you run the firmware and then reboot.

Share and Enjoy

I’m working on making some changes to the blog, mainly to make it easier to access information, so the number of pages is being reduced and the content changed to completely unlocked word and pdf documents for your download and distribution.

As part of this some of the detail we went into with regards to firmware is being scaled down to keep the content manageable for that we apologize, but accurate information is more important to us, than no information, so if it’s gone, it has gone, but the information will be directly available from the vendors’ support site.

Share and Enjoy

http://www.bladewatch.com/2011/05/05/keynote-hp-technologywork-2011-day-2-part-3/

..Photos of the presentation slides have been published here, where possible, apologies I was using my legacy iPhone 3G to take the pictures, so they’re not perfect quality, but good enough.

Coverage continues throughout the day..

We then moved on to discuss Performance Optimized Data Center, HP’s data center in a container solution, it was interesting to hear the case study discussed and how these kind of data centers are becoming increasingly seen as not only a temporary bypass solution for data center capacity issues, but also a new way of provisioning the data center as a service. Not so much a pay on use model but more one where you either buy or lease your server, the switches, the storage and the data center and at the end of the time period, you either send it back, or migrate to a replacement, love the concept.  I know there has been a lot of talk even amongst colleagues of mine about building an empty shell and deploying PODs in side it, that way deploying infrastructure on demand in volume but still giving people the re-asssurance (real or perceived) of a bricks and mortar style data center.

Key comments:

• Scales up to 3500 servers in capacity
• Supports up to 1100u
• 45% cheaper than a traditional data center
• 37% more efficient than a traditional data center

The conversation then moved on to cover the necessity to look at energy efficiency both in terms of the storage and the network.

In storage the fundamentals were:

• Thin provisioining
• Tier the storage
• Use appropriate storage platforms with data de-duplication, real time (SAN), near line (disk or NAS), offline (tape), only keeping the data that you need online or on site.

In networking the fundamentals were:

• More efficient network adapters that only use the power we need
• Convergence of the storage and network paths using wire once technology, combining storage and network virtualization where we reduce the number of powers and patch cables required to deliver the same capacity and throughpoint.

An interesting keynote, now off to the seminars and sessions..

Share and Enjoy

http://www.bladewatch.com/2011/05/05/keynote-hp-technologywork-2011-day-2-part-2/

..Photos of the presentation slides have been published here, where possible, apologies I was using my legacy iPhone 3G to take the pictures, so they’re not perfect quality, but good enough.

Coverage continues throughout the day…

We moved on to the USP offering on HP hardware and achievements in the energy efficiency space for HP servers and platforms.

• Sea of sensors – sensors built into the server which monitor and gather statistics on load and performance for analysis and management
• Energy efficient power supplies – 94% efficient power supplies reduce unnecessary utilization
• Low power memory reduces the power needed to power the server from the start
• It was mentioned how these combined with their system design, bios configuration and software tools such as:
• HP Dynamic Power Capping – which allows the server to power down components not in use, or adjust load to make the server more efficiently use the power it needs – powering down fans or individual PCI slots,
• HP Intelligent Power Discovery – identify servers are properly load balanced from a power point of view, identify power utilization and load
• HP Dynamic Power capping – allowing you to define patterns of usage and reduce power utilization on servers or server groups

All of which are extensions to the off the shelf HP Systems Insight Control Management software portfolio of products.

A question was asked regarding which products had these features included and the response that this would be discussed separately due to the volume of server models and technologies in scope for that question, fair enough.

Photos of the presentation slides have been published here, where possible, apologies I was using my legacy iPhone 3G to take the pictures, so they’re not perfect quality, but good enough.

Coverage continues throughout the day..

Share and Enjoy

The keynote continues moving on towards discussion about the Top 500 Supercomputers and Top Green 500 Supercomputer lists illustrating the achievements made by combining the right range of HP technologies to deploy not only a high performance supercomputer in the top four, but also a supercomputer that could achieve this with energy efficiency and scalability in mind powered by HP Proliant S390s G7 scale out servers.

It’s always great to highlight the opportunities that exist in deploying not only a high performance compute solution at a more accessible cost (as the cost of technologies falls) but also deploying such solutions with energy efficiency in mind through the advances in the industry such as low power memory, more efficient power supplies and better design.

The discussion quickly moved towards the Smart Grid which remains an important part of the Converged Infrastructure solution comprised of a range of technologies and services, building blocks which underpin agility and energy efficiency, building up to Converged Infrastructure and BladeSystem Matrix. Examples include:

• Sea of sensors
• Energy efficient power supplies
• Energy efficient power distribution

Photos of the presentation slides have been published here, where possible, apologies I was using my legacy iPhone 3G to take the pictures, so they’re not perfect quality, but good enough.

Coverage continues throughout the day.

Share and Enjoy

http://www.businesswire.com/news/home/20110426005820/en/PLX-Switches-Selected-Provide-Unique-x16-PCI

SUNNYVALE, Calif.–(BUSINESS WIRE)–PLX Technology (NASDAQ:PLXT), a leader in high-speed connectivity solutions for the enterprise and the home, today announced that two configurations of PLX® ExpressLane™ PCI Express® (PCIe®) switches are enabling the high-performance CPU-GPU flexible scale-out in the latest HP ProLiant SL390s G7 4U server platform.

It’s great to see further investment in the HP Proliant SL390s G7 platform, anything they can do to aid the scale out, and performance capabilities by enhancing or adding functionality in the GPU space has to be a good thing, I remain a fan of the Scale out platforms from both HP and IBM, they might not be the server solution for everyone, but for bulk system deployments like cloud, or application/grid servers, their form factor and their remote management/ease of deployment ideally suit them to scale out projects.

Share and Enjoy

We bought using the company credit card a HP MicroServer from ServersPlus.com in a bundle which included a 4GB memory chip in the box as well as a USB stick with VMware ESXi 4.1 and a DVD writer drive for £320, a virtualisation or smb bundle so it had everything you need to get started either with virtualization or as a file server.

The specifications are as follows:

Processor – AMD Athlon™ II NEO N36L
Memory support 8 GB with two slots
Expansion slots
1 half-height, half-length PCIe x16 Gen 2
1 half-height, half-length PCIe x1 Gen 2
Network Controller – 1GbE NC107i 1 Port
Storage – Maximum drive bays (4) LFF SATA
Supported drives – Non-hot plug 3.5-inch SATA
Storage Controller – Integrated 4 port SATA RAID

Also for Marjorie, our MicroServer we added as above:

4GB memory upgrade

DVD writer drive

ESXi on a USB stick

I love the HP MicroServer, I like the form factor and the fact that a remote management card is available. The specification with it’s on board RAID controller support, the four drive bays gives ample room for upgrading capacity or enabling RAID. The standard machine (the one we purchased) came with a 250GB drive which is enough to get started if we keep in mind the atypical usage scenario a file/print server, or small application or virtualization test bed.  The processor is not the fastest though adequate, 1GB RAM is getting rather low these days, but the memory can be easily upgraded and is not very expensive. Crucially though we do have to appreciate usage and the target market and features deliverable at a certain price point.

I started quite a debate with a few colleagues when asking their opinions and stating that I felt that the DVD drive should be standard, I got both responses of agreement and me missing the point completely. For the first server for that SMB using a desktop as a server absolutely migrating them, they will probably need to load their linux or Windows CD/DVD and install the operating system, but as my architect friend said, for everyone else doing VMware ESX, I have the USB on a stick I don’t want to pay for an optical drive that I probably wont use or need.

It remains a great start, it creates opportunities both as a server, as a network storage device running a range of operating systems from Microsoft Windows XP through to Microsoft Windows Server 2008 Foundation or Linux, I praise HP for the concept, for the innovation and will continue to praise the concept of a user accessible (both on price and features) server, the more entrants and innovation we have, the better as we saw recently with Fujitsu’s entrance to the market with their alternative.

With that we will award the MicroServer 8 out of 10, it does exactly what it says on the box and for many users and small businesses where a desktop is the server, this represents a great start which we here at Bladewatch are happy to recognize, well done.

Going forward for all those looking for a virtualization in a box solution, we hope and dream that version two includes maybe a more powerful processor and an optical drive? We’ll have to wait and see..

Share and Enjoy

Check out this interesting post on the IBM server blog talking about VMware on their new System X and BladeCenter servers, it tackles some of the key differentiators for running VMware on IBM’s servers, its a good read, do check it out, it’s always important to understand what others are doing and what the perceived differentiators are and why.

Share and Enjoy

So I’ve published some additional photos on our flickr page, there’ here. They include pictures of the slides from the Keynote speech from Philippe Trautmann, the HPC & POD Sales Manager for EMEA.

His introduction was regarding the state of the data center and the enterprise, what the challenges are and how they are perceived by HP and it’s customers.

Key trends:

• Government legislation on efficiency is having a big impact – for example PUE / Carbon footprint
• Small savings on energy consumption can make significant savings on operational cost
• 96% of data centers will need upgraded or modified in the next five years to keep the lights on

Coverage continues shortly.

Share and Enjoy

http://www.finextra.com/news/fullstory.aspx?newsitemid=22519

The fault is the latest in a series of banking outages to have struck the nation’s biggest banks as they rip out legacy technology and migrate to more modern core banking systems.

In a statement, Westpac group executive retail and business banking Rob Coombe says: “Westpac sincerely apologises to all our customers who have been impacted by today’s outage. We take systems reliability extremely seriously and are very disappointed by the inconvenience to our customers and will undertake a thorough review.”

Having been discussing data center optimization, energy efficiency and reducing operational costs, I saw this post on finextra talking about this bank’s outage as a result of a systems upgrade replacing their legacy technology. There’s more coverage here which is interesting, do check it out, the issue apparently was related to air conditioning.

It’s interesting to note that they are going through a refresh and migration project to realize service improvements and I suspect a reduction in operating costs. Regardless, well done to them communicating to their customers and apologizing for the loss of service.

Share and Enjoy