Home Tags Posts tagged with "Proliant"
Tag:

Proliant

ComputeDatacenterVirtualization

HP/HPE Physical to VMware conversion for Windows 2008 R2 blue screens

by Philip Sellers
written by Philip Sellers

Recently, I was converting an HP/HPE Windows 2008 R2 physical server into VMware (yes, P2V still happens), when I hit a blue screen on boot.  Turned out I was getting an Inaccessible Boot Volume stop error – a partial screenshot is below.

bluescreen 151

Problem was, the blue screen went by so quickly that I had a difficult time capturing what the stop code or issue was before the VM rebooted.  Turns out it was a simple boot issue to solve.  The VM was moving to an LSI Logic SAS controller in VMware.  That driver was not enabled in the image.  The following steps will walk you through enabling LSI SAS to allow the VM to boot.

  1. Launch recovery mode – this was automatically available on the menu after my failed boot.
  2. On the welcome screen, select Next.
  3. Launch Command Prompt.
  4. Run regedit.
  5. Select HKEY_LOCAL_MACHINE, then select File > Load Hive from Menu.
  6. Navigate to C:\Windows\System32\Config\SYSTEM.
  7. Give the Hive a temporary name.  The temporary name will be added under HKLM. I called mine TEMP.
  8. Expand your new temporary name.
  9. Find key HKEY_LOCAL_MACHINE/TEMP/System/CurrentControlSet001/Services/LSI_SAS
  10. Change Start to 0.  (mine was set to 3).
  11. Select HKEY_LOCAL_MACHINE/TEMP, then select File > Unload Hive from Menu.
  12. Reboot.

Thanks to this older post over on serverfault for the steps.

0 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenterStorage

HPE intros new ProLiant Microserver Gen10, steps backwards

by Philip Sellers
written by Philip Sellers

HPE unveiled the ProLiant MicroServer Gen10 this week at HPE Discover and while a refresh was long overdue, the MicroServer has returned to its G7 roots with the new model.  HPE decided never to release a Gen9 model of the MicroServer, though this Gen10 feels more Gen9 than new and shiny.   The model notably returns to AMD processors, switches to a SoC motherboard, and also removes the iLO from the platform.  The MicroServer was once touted for remote-office, branch-office use cases but the absence of IMPI, critical in ROBO, makes me wonder where this is targeted.

What’s Different and New?

The new MicroServer Gen10 features AMD processors, Opteron X3000 series with 2 core or 4 core as options and turbo clock speeds up to 3.4GHz.  It scales up to 32GB, double the previous Gen8 model.  The motherboard switches from a socketed Intel design to a System on a Chip design from AMD – meaning the processor cannot be field upgraded.  The new motherboard design adds two PCI slots over 1 PCI slot in the previous Gen8 and G7 models.

For storage, 4 large form factor drive SATA slots are retained in the Gen10 model and the same limit of 4 x 4TB SATA drives in the drive slot.  With SSD advancements, the new model adds the ability to replace the DVD drive with an SSD in a 5th slot near the top of the cube, but complicating things for OS installations with the absence of virtual media through iLO.

For video, the new model includes two display ports, possibly signaling the intended use case for the system around video applications – perhaps video walls and other things.  Upgraded GPU capabilities also accompany the new model, but both of these trade-off mean that MicroServer is less of a general-use server and more specifically for uses where graphics are needed – so not home-lab users or general ROBO virtualization.

HP released the MicroServer Gen8 back in 2013/2014, with a strange cubical form factor and the new Gen10 model retains this design, in fact, allowing a Gen10 to stack on a Gen8 or vice versus.  The bezel on the model now comes in a black finish instead of the default silver look of the Gen8.

HPE is also talking about ClearOS on the MicroServer Gen10.  ClearOS is a linux distribution based on CentOS and Red Hat with a marketplace of application solutions from 6 categories: Cloud, Gateway, Server, Networking, System and Reports, according to the ClearOS website.  HPE is touting it as a cost-effective alternative in the SMB market, and showed it off on the MicroServer Gen10 during HPE Discover.

Where does it miss?

With the introduction of Gen10, HPE is touting the security of their ProLiant servers – billing them as the ‘most secure industry standard server.’  With the MicroServer, the absence of iLO means the security features do not extend to the new MicroServer.  The ‘silicon root of trust’ utilizes the iLO 5 silicon to establish that NIST compliant security, so even the inclusion of an iLO 4 would not have helped this capability.

The iLO inconsistency is not a first for HPE – a couple of the lower-end Gen8 models included iLO3 chips rather than the new iLO4 that shipped with Gen8.  Some of the Gen9 models are missing an iLO all together.  The compromise is certainly targeted towards bringing costs down, but its a trade off that I’m sad to see.  Even the original MicroServer G7 model had the option of an remote console card in the PCI slot – a waste of a PCI slot in my opinion – but at least it had the option for anyone who found it important.

If HPE wants to follow a SoC motherboard model for this line, I hope to see a Xeon-D variant in the future. A socketed model is better, but with a wide support of the Intel SoC models, I hope that at least we get an option in the future.

I’ve used a ProLiant MicroServer Gen8 in my home lab for years.  I sold it during my move, but it is truly one of the best home lab models I have found, mostly because I prefer a strong and capable IPMI.  So, that’s why I think I’m taking the absence of iLO so hard in this new model.  The community waited for a long time for this and the home lab options from SuperMicro are abundant, even with a sub-par IMPI interface.

8 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenterNetworking

With new HPE EdgeLine systems, its about data proximity

by Philip Sellers
written by Philip Sellers

Less than a year ago, HPE announced a new set of products under its EdgeLine banner.  These devices were created to enable the Internet of Things (IoT) to connect with other IT infrastructure and to allow IT to tap into sensor data and other capabilities directly attached to devices.   With the term, Internet of Things, visions of connected refrigerators and connected toaster ovens spring to mind, but when it comes to the Enterprise, where the new HPE is clearly focused, what does an IoT play look like?

HPE believes that when it comes to IoT in the enterprise, it is all about empowering processing and analytics to occur close to where the data is being generated.  This past week at HPE Discover in Las Vegas, HPE announced the EdgeLine 1000 and EdgeLine 4000 series of devices that allow more processing power to be deployed into the field near these connected devices that generate data.  These devices join the EdgeLine EL10 and EL20 gateway devices it introduced in December at HPE Discover in London.  The idea is that by processing the data in the field, the enterprise can save on data transmission, security and can really do some interesting new things in the field.  HPE sees this as critical in the success of IoT for enterprise.

The EdgeLine 1000 uses a single ProLiant cartridge (the kind used in HPE Moonshot) in a ruggedized metal case with built-in expansion slots for drives and PCI cards.  The EdgeLine 1000 shown off at HPE Discover also featured integrated wireless networking on the system.  The EdgeLine 4000 accommodates up to 4 ProLiant cartridges and resembles a traditional rack-mount server.

Why Process Near the Data?

Many of these industrial uses of IoT may be in remote and disconnected parts of the world.  It may not be feasible or cost-effective to pay for the data plans required to connect these devices and push all of that data to a datacenter or the cloud.  Remote oil platforms, refineries, industrial sites, and even remote jungle deployments [like the one HP showcased with Conservation International] may not have available connectivity options.  Or if they do, satellite internet connections are metered and expensive.  And even in well connected areas, it may come down to security and not wanting these devices actually talking to the Internet.  There may be regulations or laws that govern what connectivity can be established – or it may be simple corporate secrets where they do not want to introduce the risk of transmitting the data outside of its corporate IT.

While at HPE Discover, the team from DS Virgin Formula E racing talked with our group of bloggers.  One of the most interesting things about Formula E racing is the amount of sensor data created by the electric race cars and then the amount of data analytics that the teams are performing on that data in the field.  For the DS Virgin Formula E team, this was a huge contributor to a much more successful season in year 2.  During the first year, the team did not have analytics in the field and were purely racing based on instinct and some review of the sensor data manually.  Now the team has deployed HPE Moonshot in the pit and they can download and analyze the sensor data received from the car.  In a typical race day, the team comes to a track, has no previous experience with the track, and after a couple rounds of qualifying with short times in between, they run the final race.  By using analytics, they take the data obtained by the car in qualifying and they can make adjustments and plans about how to run the race best and most efficiently.  The results show for themselves with the team taking the poll in most of the races this season and ranking currently 3rd in the entire field for the year.  Taking the same data and sending to to the cloud is a waste of time – the transmission time is better spent doing analytics during very tight turn-arounds.  The same principal applies with IoT data.

You also have to determine what data is actually valuable.  Everyone talks about the astronomical size of the data that will be generated from IoT devices.  Even if you have good connectivity, using all of your available upstream bandwidth to transmit sensor data is probably not the best use case of your connectivity.  Case in point, I personally started with a DropCam at home for security reasons – but I dislike the device because it clogs my upstream – constantly transmitting video the cloud.  Instead, I have setup another set of cameras that only transmit minute long clips when a camera detects motion.  That’s really what I care about – knowing that something moved and seeing video of that – not broadcasting 24/7.   IoT generated data is a lot the same and this is where HPE’s vision lies – putting the software to detect the motion or a failure or other event near the data generating device and then allow the detection to occur there.

Unlocking New Capabilities

Also by having processing available in the field, you open up new capabilities to remote workers.  HPE was demoing a pump system from FlowServe on the show floor at HPE Discover.  FlowServe makes industrial pumps for major operations around the world.  Approximately 10,000 of those pumps are outfitted with sensors that can communicate problems, status and operational data points.  FlowServe has been working with HPE and the EdgeLine product line to take their sensor-enabled pumps and extend those capabilities by connecting them.

Ckin_ewVEAEaZEAIn the demo I saw, FlowServe was able to connect the pump system to an EdgeLine 1000 system and then run real time monitoring of the pump.  With these located thousands of miles away from main office locations, this sensor data could enable operators to know when conditions that may ruin a pump occur and shutdown the system to prevent problems.  Now, that’s not new – what is new about the solution is that the EdgeLine system could then transmit that problem to a help desk and arrange for a repair automatically.  That’s the power of connectivity.

Once a repairman is onsite, however, he may have never seen the installation and may not know anything about where the actual problem occurred.  HPE and FlowServe were showing off an augmented reality program that paired an iPad with special software that take a live picture of the installation and overlays sensors and status to the live picture.  This augmented reality app can show the repairman where the problem has occurred, give them important data about what needs to be services and even allow the repairman to start the system back up and ensure that his repair was successful by communicating security with the EdgeLine system in the field.  A photo of the augmented reality app is to the right.

Smart Cities

HPE was talking specifically about industrial applications – manufacturing, oil, gas – but an interesting fourth example – smart cities.  In the case of smart cities, HPE is thinking about traffic cameras, environmental sensors and other applications of technology to connected devices.  In its concept of smart cities, HPE showed demos of connect vehicles communicating with one another or with drivers to warn about potholes or other obstructions in the roadway – sort of like Waze, but with much more automation and precision.  Booths also showed off solutions for smart parking, health care, supply chain, waste management, the connected car and personal safety tracking devices.  All of the solutions were built around interactions that we would have with technology to assist us in our daily lives.

I also talked with a security researcher in Hewlett Packard Labs and we had an interesting discussion about pushing facial recognition analytics out to individual cameras where specific persons of interest could be recognized by the camera, rather than having to transmit all the video data back for centralized analytics to run on it.  Innovation like that could enable quicker arrests and pinpointing terrorists before they strike.

While I walked around the booths demoing concepts for the smart city, I have to admit that I thought about TomorrowLand and Walt Disney.  This had to be sort of how TomorrowLand felt when it first debuted.  Very futuristic concepts of how a smart city could run.

0 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenterSoftwareStorage

Simple iLO Firmware updates with PowerShell

by Philip Sellers
written by Philip Sellers

ilo

The Hewlett Packard Enterprise iLO PowerShell cmdlets make it extremely easy to update iLO Firmware in mass.    The Update-HPiLOFirmware cmdlet is an extremely handy for updating firmware across many systems.  I particularly like this method because it fast, which isn’t always the case when trying to script upgrades.  The syntax of the cmdlet is easy to remember and with Get-Help not far away, even if you forget the simple syntax, you can easily find the details.  Combine the Update-HPiLOFimware functionality with the Find-HPiLO cmdlet used to discover and report iLOs using a network range and you have a winning combination to automate iLO firmware.

This method proved to be handy for me because I make a point to update iLO firmware prior to updating servers with a new Service Pack for ProLiant (SPP).   Since I utilize the virtual media feature in iLO to deploy the SPP to ESXi nodes, I have found it is best to update iLO first and then let the SPP install all other firmware.  If the iLO firmware is updated and it restarts during the SPP run, this causes the installer to fail because it lost its media.

Before you start

There are a couple quick things you’ll want to ensure before you start.

  1. You’ll need the HP iLO Toolkit.  It is available here – http://www8.hp.com/us/en/products/server-software/product-detail.html?oid=5440657.
  2. Make sure you have a consistent admin credential for iLO.  It makes updating multiple hosts much easier.  The credential can be a local account in iLO or it can be an account from Active Directory if you have defined and setup AD integration on your iLOs.
  3. It is easiest to pass your credentials by variable using a PowerShell credentials object.  You can do that easily with the Get-Credential cmdlet.

    [code lang=”Powershell”]$ilocreds = Get-Credential[/code]

    Alternatively, you can use the -username and -password switches on the Update-HPiLOFirmware cmdlet later.

Step-by-step

Working through the process step-by-step, you’d run the following commands to update firmware:

  1. The first step is to get a list of the servers using the Find-HPiLO cmdlet.

    [code lang=”Powershell”]$servers = Find-HPiLO 192.168.200.1-255[/code]

  2. Next, you can investigate the current version of firmware on each iLO found with this simple Select and Sort statement.

    [code lang=”Powershell”]$servers | Select HOSTNAME, PN, FWRI | Sort PN, FWRI [/code]

  3. Now that you’ve seen what the Find-HPiLO cmdlet returned, you can scope down the list to what you want to update.  In this example, we are looking for firmware older than version 2.22 on iLO4 in teh Where statement.  Confirm the output list of servers are the servers you wish to patch.

    [code lang=”Powershell”]$servers | Where {$_.FWRI -lt 2.22 -AND $_.PN -like "*iLO 4*"} [/code]

  4. Repeat the scoped list and now pipe that into the Update-HPiLOFirmware cmdlet.  This cmdlet need the credentials you setup earlier and it accepts the list of servers through the pipeline.  The last switch needed is the location of the firmware file to load.

    [code lang=”Powershell”]$servers | Where {$_.FWRI -lt 2.22 -AND $_.PN -like "*iLO 4*"} | Update-HPiLOFirmware -Credential $ilocreds -Location C:\firmware\ilo4\ilo4_222.bin[/code]

Condense it down

The same can be done as a one liner too – Find, Where (scope) and Update…  You still need to set your credentials ahead of time.

[code lang=”Powershell”]Find-HPiLO 192.168.200.1-255 | Where {$_.FWRI -lt 2.22 -AND $_.PN -like "*iLO 4*"} | Update-HPiLOFirmware -Credential $ilocreds -Location C:\firmware\ilo4\ilo4_222.bin[/code]

HP PowerShell Toolkits

In addition to the iLO PowerShell toolkit, HPE also has sets of cmdlets to administer BladeSystem OnBoard Administrators and HP ProLiant BIOS settings.  You may grab all three sets of cmdlets and find full documentation at http://www8.hp.com/us/en/products/server-software/product-detail.html?oid=5440657.

1 comment
0 FacebookTwitterPinterestLinkedinRedditEmail
DatacenterSoftwareStorage

Deploying HP OneView 1.20 for monitoring & replacing HP SIM

by Philip Sellers
written by Philip Sellers

oneviewheaderWith the 1.20 release of HP OneView, HP has opened up the ability to monitor hardware without controlling all configuration and deployments using the tool.  Along with the 1.20 release, HP introduced the OneView Standard license, which is a free license that allows HP customers to run OneView for monitoring only of hardware.  This license level lets users use the dashboard  and alerting features to view all of their ProLiant G6 and G7 hardware along with Gen8 and Gen9, including blade enclosures.   These changes allow OneView to be a drop-in replacement for system management and replace HP System Insight Manager in many deployments.

For comparison, with the full-featured OneView Advanced license, customers can manage deployments on ProLiant Gen8 and Gen9 hardware.  Customers can create server profiles and provision these onto blade and rackmount ProLiant hardware.  In addition to ProLiant, OneView is also able to control network configuration for blade hardware and storage configurations connected to blade and rackmount hardware.

Hands-On Impressions

For anyone reading who may be accustomed to HP System Insight Manager (SIM), HP OneView is a huge leap forward.  The UI is intuitive and easy to learn.  The Standard license level allows most people to replace HP SIM if they have current generations of ProLiant hardware.

Once I deployed HP OneView in my environment, all that was needed was connecting it to my blade enclosures and ProLiant rackmount servers.  The connection point is the iLO, so you fill in the iLO’s information, a username and password and then HP OneView configures its own user account on each server to connect and poll from the system.

After connecting the systems, you have a Dashboard view that shows you system health and problems.  Its a pie chart that will show you color coded sections for systems either incompatible (in grey), healthy (green) or with warning and problems (yellow and red).  You are able to click and drill down into each one of the slices to get filtered information about warnings or errors.  The inventory is great and has good controls for filtering by state.  In addition, you can look at servers by hardware type and see how many of each model you have plus, how many different configurations you have within a particular model – based on expansion cards and stats.

Admittedly, there is a lot of additional functionality that remains empty in a monitoring only configuration.  Server profiles and interconnects areas will be blank.

In comparison with HP SIM, I’ve had OneView running for monitoring for months and have had to do zero babysitting.  The alerts that come from OneView are superior to SIM.  The alerts correlate the system name and alert state correctly, where SIM often has difficulty identifying the system name and usually reports things under the system serial number instead.   We have them running in parallel currently and I get alerts from both systems, but I get more useful information from the HP OneView alerts.

I’m not sure I know anyone who would willingly choose HP SIM over HP OneView after giving a test drive.  SIM still has a place if you have a lot of legacy hardware, ProLiant G5 and below.  But, so far, this is a great step forward for system management as an HP customer.

FAQ for HP OneView 1.20 Release

0 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenterStorage

Installing Windows 2012 R2 on ProLiant DL380 G7 without SmartStart

by Philip Sellers
written by Philip Sellers

Since HP SmartStart CD went the way of the dodo bird before Windows 2012 R2 was released, HP ProLiants older than a Gen8 require administrators to manually install Windows 2012 R2.  On ProLiant Gen8 and higher, you have Intelligent Provisioning – the replacement for SmartStart – embedded on the system, but this isn’t available on G7, so the user needs to manually install Windows according to the “Implementing Microsoft Windows Server 2012” technical whitepaper.

error

In the case of the HP ProLiant DL380 G7, the normal Windows installation doesn’t recognize the P410i storage controller during installation and it fails with the strange “We can’t create a new partition or locate an existing one.”  Searching forums, you find several solutions suggesting boot order or removable mdia like USB or SD storage causes the problem in many cases, but for the DL380 G7, however to get past this on my servers, I needed to download the storage controller driver and add it during installation.

You will need a USB drive where you can load the drivers during Windows installation.

The steps to do this and successfully install Windows 2012 R2 are below:

  1. Browse out to hp.com and go to the Drivers & Software download section or directly to http://www8.hp.com/us/en/drivers.html.
  2. Search for the DL380 G7 in the search and click on the link for the server displayed after search.
  3. Select Windows 2012 R2 as the operating system dropdown.
  4. Navigate to Storage Controller and download the latest “HP ProLiant Smart Array SAS/SATA Controller Driver for Windows Server 2012 x64 Edition.”  If you use this link, make sure to check revision history to ensure this is the latest available.
  5. Click the Download button and then run the .exe file.  Use the Extract button to copy all the files onto a USB stick.
    c03932407
  6. Boot the system and begin Windows 2012 R2 installation from CD or from ISO connected via remote console on the iLO.
  7. Click the Install Windows.
  8. Custom Installation.
  9. Storage window – choose Have Disk.  Make sure USB is plugged in on the server.
  10. Choose the HP driver
  11. Click next and the installation will now continue instead of giving the error.
    WinSetup
12 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenter

Adjusting UEFI boot order on ProLiant DL580 Gen8

by Philip Sellers
written by Philip Sellers

When HP introduced the ProLiant DL580 Gen8 last February, HP also launched it’s new UEFI firmware, which replaced traditional BIOS and became the standard across all ProLiant Gen9 servers introduced last September.  Typically, the powerhouse ProLiant DL580 servers lag behind the introduction of other generational servers and the Gen8 model was no different, being introduced well over a year after Gen8 debuted.  Its pretty safe to assume that HP sells many less units of the DL580 model than other ProLiant rackmounts and it was a pretty smart strategic move to add UEFI to this model before launching it across the entire line of mass-market servers.

Some of the menus in the UEFI inteface on the ProLiant DL580 Gen8 are a little clunky.  One of those is the boot order screen.  After several failed attempts at changing the boot order, I engaged the help of a product manager within HP who explained where I was going wrong.  And for the benefit of others, I wanted to put this out to explain how to make boot order changes on the ProLiant DL580 Gen8.  The trickiness comes when you actually make the change – getting to the boot options is straight forward.

Fortunately, the same product manager confirmed that ProLiant Gen9 servers have an improved version with clearer instructions.

Changing Boot Order on a DL580 Gen8 with UEFI

During the normal boot sequence, you get the normal POST screens and options along the bottom – one of those is F9 for System Utilities and you should press F9 to invoke those menus.

FmenuOptions

Once the System Utilities menu appears, choose the System Configuration option and press enter.

systemConfig

From System Configuration, you will choose Boot Options and press enter.

bootOptions

From the Boot Options, you come to the RBSU and the first option is the UEFI Boot Order.  Choose UEFI Boot Order and press enter.

This screen also confirms whether you are using UEFI Mode or Legacy BIOS for booting.  If you wanted to make a change, this is the screen where you change to Legacy BIOS from UEFI or vice versus.  If you make a boot mode change, all options will get greyed out and you will need to reboot and go back into System Utilities to make further changes.  

rbsuMenu

Once you’re in the UEFI Boot Order screen, you see a list of options along with brief instructions.

bootOrderScreen

You must press enter to go into the boot order screen and make changes.  Scroll to the item you want to move in boot order and press the + or – keys to move the item up or down the list.

makeChanges

Once you make the changes, press Enter to save those changes.  There is no onscreen prompt that says this, but pressing F10, even though it says Save in the display, will not save it.  Pressing ESC will lose your changes.

savedChanges

Once you press Enter, you go back to a screen and should see your new boot order confirmed.  You may press F10 here to save the settings.  Press Enter as instructed on the on-screen prompt and this saves the changes.

pressF10

When the system reboots, now you should immediately begin booting from your USB device or hard drive that you moved to the first position in boot order.

1 comment
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenter

ProLiant servers getting RESTful API for management

by Philip Sellers
written by Philip Sellers

ProLiant servers are getting some REST – a RESTful API.  One of the innovations HP is showing off on the HP Discover showroom floor is a new RESTful interface packaged as part of the iLO management chips on ProLiant servers.  With the RESTful API, HP hopes to deliver future generations of ProLiant that can be easily provisioned and controlled from tools like HP OneView.  RESTful interfaces are becoming the go-to standard within HP for interoperability between their products.

HP RESTful API is a management interface can be used to perform configuration, inventory, and monitoring.  The RESTful API is available on Gen8 and Gen9 hardware with iLO firmware 2.0 and greater.  On Gen9 and future hardware, the interface will be able to configure both BIOS and iLO settings on the server.  On Gen8 servers, the RESTful API will only expose the capabilities inside of the iLO.

According to HP, a REST client will send HTTPS operations to the iLO web server and there are three modes that can be used – command line interface, file based and JSON-formatted data.  The use of HTTPS is a big security plus versus many other remote management interfaces that are often

The folks at HP are working with other vendors in the market to establish a standard that they call the Redfish Specification.  Redfish is supposed to create a standardized interface for IP Management Interfaces on servers to allow uniform control and reporting across multi-vendor server environments. The standardization should allow the same management software to be able to manage heterogeneous servers.  Redfish is corporately sponsored by Dell, Intel, HP and Emerson.

For more information, see http://www.hp.com/go/restfulapi.

Disclaimer: HP invited and paid for my trip to attend HP Discover EMEA.  The views and opinions I’ve written were not dictated or approved by HP and are my own.  

0 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenterVirtualization

Stick to the recipe to keep ESXi on HP hardware cooking

by Philip Sellers
written by Philip Sellers

Back in September, I blogged about a potential problem between HP’s September software bundle for ESXi and ProLiant Gen8 blade servers with Emulex CNA’s.  Long story short, the issue was caused by a mismatch of firmware between Virtual Connect and the Emulex CNA’s.  The mismatch was induced by me, as the administrator, not keeping all my components on a tested firmware ‘recipe’ from HP.

A couple years ago, HP made a shift in the BladeSystem firmware that unified the individual components into a tested configuration and began publishing the firmware together as release sets.  This meant that if you installed a recipe of Onboard Administrator, Virtual Connect and iLO firmwares, they would be more stable since they had been tested together for interoperability.  About a year later, HP unified software drivers and firmware on all ProLiant systems in a similiar way with the release of the Service Pack for Proliant (SPP).  The Service Pack for ProLiant did the same thing for OS level drivers and hardware firmware that the earlier release sets did for BladeSystem firmware.  All components were tested together at specific levels and the SPP’s unified release made it easy for administrators to apply a blessed recipe of software across components.

HP also does a similar thing with its OEM version of ESXi, which includes additional software components that allow a Common Information Model (CIM) implementation within the OS to display hardware status among other things.  Because of the add-in software, HP maintains a recipe between this software and the SPP bundle and firmware levels required and tested on their hardware platforms.

The moral of the story is to stay with HP’s recipes and SPP releases.  When you deviate from these, having firmware older than recommended along with newer drivers or agents, problems can occur.  In my own experiences, when I have stayed with the recipe, I haven’t experienced major malfunctions.  As with any software, new releases fix different bugs and there are times when you want to move outside of a recipe, if a newer version fixes a specific issue you’re experiencing, but in general, the rule is stay in recipe.

The other great benefit that you receive as a customer is a guarantee of support.  Traditionally, HP support pointed to older firmware as the first course of action for many support calls.  Following the SPP and ESXi recipe models, you now get a year of support from the release date of a bundle.  With this, you can push back on support and say “I’m in a supported configuration, now support me.” if you’re ever asked to update firmware without troubleshooting first.  This is a great position for the customer.  Support can also point to specific resolutions in release notes if there is a problem resolved that requires you to move to newer release, but following HP’s guidelines ensures that all the separate components have been tested together for the best possible operational environment that you can have.

And that advice is based on years of struggle in the pre-release set world of HP hardware, experience with HP support and my own personal lapses in following my own advice…  Hope this helps you avoid my struggles…

Service Pack for ProLiant – http://www.hp.com/go/spp

VIBS Depot from HP – http://vibsdepot.hp.com

2 comments
0 FacebookTwitterPinterestLinkedinRedditEmail
ComputeDatacenterSoftwareStorageVirtualization

HP announces Hyper-Converged Systems based on StoreVirtual VSA

by Philip Sellers
written by Philip Sellers

HP unveiled the HP Converged System 200-HC StoreVirtual based on HP ProLiant SL2500 series of scalable computing servers and the HP StoreVirtual VSA today at VMworld in Barcelona.  The CS 200-HC StoreVirtual features VMware vSphere 5, the HP StoreVirtual VSA, and HP OneView for management.  The CS 200-HC StoreVirtual line comes in two different models – one with all SAS for higher capacity and one with a mix of SSD and SAS drives for better performance.

Four nodes of compute fit into just 2U of rack space, with shared redundant power for the unit.  From a physical size standpoint, these nodes fit somewhere between a Moonshot card and a BladeSystem blade in terms of sizing.  While they more closely resemble a blade, there is no shared interconnect back-plane.  Each node comes with dedicated 10Gb Ethernet and 1Gb Ethernet network ports and has 6 dedicated disks.

A new addition to the software mix for the Converged System 200-HC line is HP OneView InstantOn  – a streamlined deployment utility that lets customers spin up vSphere with StoreVirtual VSA in a matter of minutes.  According to HP, all that a customer needs to provide is IP addresses and basic license information in order to have a fully working system delivered directly from the factory.

CS 200-HC

A comparison of the hardware configuration shows just a few differences between the two initial StoreVirtual models:

CS 240-HC StoreVirtual System

  • 4 nodes, each with
  • 2 sockets, 8 cores each socket at 2.0Ghz
  • 128GB of RAM each node
  • 2 x 10GbE and 2 x 1GbE NICs per node
  • 6 x 1.2TB SAS SFF drives per node

CS 242-HC StoreVirtual System

  • 4 nodes, each with
  • 2 sockets, 10 cores each socket at 2.8GHz
  • 256GB of RAM each node
  • 2 x 10GbE and 2 x 1GbE NICs per node
  • 4 x 1.2TB SAS SFF drives per node
  • 2 x 400GB SSDs per node

In terms of scale, up to 8 of the CS 240-HC line units can be coupled together to form up to 32-node clusters and not all of those nodes need to be in a single site because of StoreVirtual’s native replication and multi-site capabilities.

StoreVirtual VSA features Autonomic Optimization, a fancy name for storage sub-lun tiering that the CS 242-HC StoreVirtual is using.  This allows for more often accessed blocks of data to be moved from SAS onto SSD and less used to be moved from SSD down to SAS yeilding greater performance from the storage subsystem.  StoreVirtual also provides a wide range of interoperability options for customers who have already invested in dedicated StoreVirtual 4000 hardware solutions or who have existing StoreVirtual VSA deployments.  In addition, native replication is available to any other StoreVirtual, regardless of form factor, and StoreVirtual offers support for vSphere Metro Storage Clusters, offering the ability to create multi-datacenter replication solutions around the technology.  Beyond native replication, HP StoreVirtual also includes Peer Motion capabilities to allow movement of data from one StoreVirtual environment to another for easy migration.

The advantage to moving to a Converged System solution, according to HP, is easier management and a better support experience for the solution.  The solution is tested and supported together as a package, offering customers a single point of contact for their support.  The firmware, drivers and software are all tested together and the will receive recipe updates so that known good combinations of the software can be run on the platform.

Choice is always at the forefront of HP’s strategy for customers and the hyper-converged system announcement is no different.  The HP Converged System 200-HC models are available in both a StoreVirtual VSA and EVO:RAIL configuration.  The primary difference between the two configuration is the virtual storage appliance, with the EVO:RAIL configuration coming with VMware’s vSAN, though I cannot find any specifics of the EVO:RAIL hardware configuration at this time.

When I was in Houston, TX, for a Converged Systems Tech Day last month, HP staff talked with me about the overall Converged System strategy and there were several key points to the overall strategy.  With all Converged Systems, there is an accelerated deployment timeline with the solution.  Since it has been tested and integrated, the product should deploy quickly so that customers realize faster time to value.  The second point is easy procurement, with a single SKU for the entire solution.  The single SKU also provides single solution support, with a specialized team that supports the entire solution rather than the individual components.  Customers who have built their own solutions from ProLiant and HP Storage have probably experienced the piece-meal support experience, even while in a single vendor. HP’s Converged Systems eliminates that problem by sending all support to a dedicated support team to team to troubleshoot problems and that team understands the entire solution, not just an individual component.

The new Converged System 200-HC StoreVirtual models will be available for purchase in December.  The Converged System 200-HC EVO:RAIL configuration will be available first quarter of 2015.

For more information, check out Around the Storage Block post and video on the CS 200-HC StoreVirtual solution and go to www.hp.com/go/hyper-converge.

1 comment
0 FacebookTwitterPinterestLinkedinRedditEmail
@2019 - All Right Reserved. Designed and Developed by PenciDesign