content top

Transform, Digitise, Agile, Cost Reduction and more with less...

I was speaking with a consultant friend of mine about transformation and he was asking about the determinants of success, this was following on from a conversation that I had in which you might deliver a project but fail to meet your key performance indicators – your determinants for success. It stems from the following transformation means many things to many people, key performance metrics can mean many things to different stakeholders and indeed success can be internalised within a business line but not the wider group.  Transformation could be me migrating an application from a legacy data warehouse platform on an unsupported version of Solaris to a new Big Data solution, but often in such scenarios the financial savings are not direct in nature, we take out a suite of legacy costs (which should not be discounted) and introduce a different range of services or cost resulting from the replacement Big Data solution. Therefore defining what your key performance metrics are and what those mean is crucial. In the example above to HR, I might have reduced their spend on Solaris by $40,000 a year in IT charges, to the Solaris team, I might have enabled a physical server and some associated costs to be cancelled, but to group IT, the costs might be neutral. The level of cost transparency you have not only in the re-charge but the direct PnL charges need to be in place so that at the right level you can align projects, and goals with outcomes, a financial SQL query if you like, to achieve this target what do I need to transform or take out.  At the same time we need to recognise that decommissioning a platform, application or service should create opportunities for savings as we scale up those changes. If we transform enough legacy platforms, for example we got rid of AIX or Solaris entirely from the estate does that enable the move to Cloud more transparently with the associated opportunities to reduce costs across the stack. Transformation, the move to Digital adoption of DevOps and Agile are all admirable activities where the legacy IT consumers should be headed often the drive is to maximise change for lower direct initial costs but that can have unintended costs either from outages, unplanned behaviour or simply things like the teams not being...

HP Enterprise Discover 2016 London

https://www.hpe.com/events/discover Discover 2016 is in London next week. There is full coverage at www.hpe.com/discover which I will be sure to be checking out, and pending my schedule I may well try and turn up on the day.

HPE Intelligent Provisioning overview

There are two versions of HPE Intelligent Provisioning, for HP ProLiant servers dependent on the generation or ‘age’ of the machine: Version 2.50 supports all new HPE ProLiant Gen9 platforms. Version 1.63 supports all new HPE ProLiant Gen8 platforms. You can upgrade the version of HPE Intelligent Provisioning without affecting data, this is done by downloading the ISO and booting from the ISO, it runs the update and reboots. It lets you: Perfom an assisted OS install (and configure the array if necessary) Run maintenance/diagnostics The steps therefore for a new server build out would be: Configure ILO Download the latest appropriate Intelligent Provisioining ISO and run the update Download and run the latest Service Pack for ProLiant (currently April 2015) Run Service Pack for ProLiant and bring the server firmware up to date Boot the server and run the Array configuration utility F5 – configure the array, set the boot order and reboot Run F10 to boot into Intelligent Provisioning and perform OS install – select custom install to specify the server name and specifics (including drivers/management software if you want) If the install fails due to language settings try leaving language US English Complete OS install and configuration Re-run the Service Pack for ProLiant to install any missing...

P2v quick reference guide

I was speaking with a colleague about some work I’d been doing with a colleague about p2v (converting their physical servers to virtual machines).  In this case it was a mixture of Compaq and IBM servers running Windows 2003/2008 which they needed to migrate to virtual machines to reclaim data center space whilst the project to upgrade to Windows 2008/2012 was initiated. He asked how I approached the p2v process and I’ve made some notes below. Check ILO/IMM works ok and that you know the local administrator password – for post conversion/roll back Verify server health (nothing is alerting that might potentially stop the server from booting) Delete temporary files if possible – c:\windows\temp and c:\temp also IIS logs/middleware logs Defrag the hard drives beforehand – you don’t have too but it’s nice to do and shouldn’t take long Set the hardware management agents and non essential services to manual ipconfig /all >c:\ipconfig.txt – Captures networking information to a text file. Download pspcleaner referenced here. At this point some fundamentals of approach.  I always wanted to do the conversion in as few reboots as possible, so at a high level it’s the following: Before P2V steps: Disable server monitoring including hardware monitors/backups Log on as local administrator (verifying password) Verify accessiblility of IMM/ILO (use online tools to reset password if required) Install VMware vCenter Converter Standalone Set HP/IBM hardware agents to manual Clear down space – log files/temporary files etc Using command prompt: ipconfig /all >c:\ipconfigbfc.txt Defrag server – defrag c: -a -v defrag d: -a -v Set the flag (converter-worker.xml) to ssl value = false Restart VMware services Run conversion Conversion steps Remove networks – we will add network cards later Advanced Options – Select power on destination machine Advanced Options – Select install Vmware Tools on the destination virtual machine On completion of the P2V shut down the physical server and start a ping against primary network card Post conversion steps Log in to the server locally using administrator account Say no to the message apply settings and reboot In vSphere system settings add network card(s) and remove audio/usb Select edit settings and options, click VMware tools now click enable ‘Check and upgrade VMtools during power cycle’ Click ok and shut down the virtual machines Power on the server – if VMware tools is out of date...

Infrastructure transformation thoughts

I was speaking with a colleague asking about how I mean about infrastructure transformation and how I approach change/transformation. Infrastructure transformation means different things to different target audiences, if we take the infrastructure or IT stance. I’m talking about change which is significant enough to deliver enhancement and enablement to the end user community typically with some outages or service disruption. My approach has always been the baby step approach, fix the fundamentals and everything else follows: Analyse the infrastructure and businesses in scope Identify ideal platforms that can be test cases for transformation Command and control the transformation or change Understand our successes and what could be done better which should feed back into analyse the infrastructure in scope Command and control is something that is very much to my heart and is stated in terms of communication and methodology: A unified line of communication throughout the transformation process; outages, failures are all controlled, declared and resolved transparently. Involve the stakeholders and clarify roles and responsibilities – do it right, do it once Clarity in how we ‘undo’ the change/transformation Red/yellow/green so we have a dashboard reference point of what is or is not complete In summary: Create ocumented and repeatable steps which are easily replicated, that are transparent and includes no assumptions. Avoid ackronyms and complexity where possible to embed civility  and partnering in...

« Older Entries