May 24

Veeam Backup & Replication 5.0.2 release

I’m a bit behind, but Veeam have released version 5.0.2 of its rather excellent Veeam Backup & Replication product. Not a major release, and some minor new features, but it does fix some outstanding issues its had:

  • Internet access from virtual lab. SureBackup virtual labs now provide an optional internet access from fenced-off environment. This enables using On-Demand Sandbox for additional use cases, such as more streamlined patch or software upgrade testing. Internet access is disabled by default, and must be configured through new registry parameters on the Veeam B&R server.
  • Automated retention for removed and deleted VMs. Data belonging to removed or deleted VMs will now be automatically cleaned up from backup files according to new UI retention setting in days. Previously, it was necessary to remove such VMs from backup manually to explicitly confirm that VM data is no longer needed
  • Optional catalog import. Guest file system catalog import was made optional to reduce the time required to import backup in cases when importing catalog data is not required. The corresponding option has been added to the Import Backup dialog.
  • Date Modified attribute preservation. Instant file level restore for Microsoft Windows guests now preserves Date Modified attribute on all restored files
  • Microsoft Windows Storage Server support. You can now install backup server and Enterprise Manager servers directly on your Microsoft Windows Storage Server.

May 24

Starwind – Free ISCSI SAN

Starwind have recently released a free version of is ISCSI SAN product set, with no capacity limits!

It includes the following features:

  • Data Deduplication: Data Deduplication is a specialized mechanism of data-compression that allows reducing needed storage capacity due to the process of eliminating duplicated data
  • Snapshots & Clones: Capture point-in-time snapshots with unlimited rollback points or make full volume copies
  • CDP (Continuous Data Protection): CDP provides a replication mechanism to ensure an RPO (Recovery Point Objective) of zero data loss for applications that require a lower RPO than asynchronous replication can provide
  • Thin Provisioning: Allocate space dynamically for highly efficient disk utilization.
  • Support for iSCSI Standards: IPSec, CHAP, MPIO are all fully supported in StarWind Free.

Starwind has many more features compared to that of the Microsoft iSCSI offering. It is, for example ideal for home labs. Just provide a computer with a bunch of cheap drives and this can be transformed into a shared storage device, with some nice almost enterprise level features.  This lets you test Live Migrations, vMotion, backup solutions etc.

For more details: http://www.starwindsoftware.com/starwind-free

Apr 19

First EMEA HP Channel Partner Matrix InstallFest

Being described as elite and crème de la crème is always a brilliant way to start a course.  But this was no ordinary course I was attending, because of ITEX’s (the HP Channel Partner I currently work for) outstanding partnership with HP, and coupled with my not too shabby qualifications 🙂 I had been invited to attend the very first EMEA HP Channel Partner Matrix InstallFest in Grenoble.  The InstallFest is not a training course, it’s a full assessment of the ability of the consultant to deploy the HP Bladesystem Matrix (HP’s enterprise public & private cloud offering) in their first steps towards becoming a Master Matrix Installer.  At the moment, there are only 85 Master Matrix Installers in the world, and they all come from within HP, a select number of HP Channel Partners have been invited to participate in creating new Master Matrix Installers from within their ranks, of which, I am working towards becoming one.

So, what did this InstallFest entail?  5 days of constant assessment by 2 Master Matrix Installers who both have dozens of installs under their belts and who have contributed to the actual design of the HP Bladesystem Matrix solution.  I was tasked with deploying a solution as would be completed at a client site and every stage of this was commented on and evaluated by the assessors.  It might sound easy, but there is a lot of pressure to prove yourself, and considering a Matrix solution comprises of virtualization products, storage area networks, virtual connect networking, blade based hardware and a plethora of HP software products, all of which need to be configured in the correct way such that this enables them to interact and communicate with each other correctly, its very taxing on the brain.

It wasn’t all work work work though, we did enjoy the sights of Grenoble, and had some pretty interesting meals out, highlights including the South African team drinking a restaurant dry of beer and the Spanish contingent annoying the waitress by trying to take photos of her all the time! 🙂

So, after all the hard work, and a little fun :), what was the outcome? I graduated!  I am now part of the mentoring scheme, which means I have to now complete two live Matrix client deployments alongside a current Master Matrix Installer, once these are completed and assessed then the worldwide Matrix solution team meet to decide my fate, and if they agree I will be granted the Master Matrix Installer role within HP and the Channel Partner program, joining a rather select community of consultants across the world!  Watch this space for my progress…

Apr 18

To FOM or not to FOM, that is the question!

You have a mutli node P4000 environment, do you need a Fail Over Manager (FOM) or not?

But first, what the hell is a FOM?

The FOM is a specialized version of the SAN/iQ software. It runs as a virtual appliance in VMware that can be run on ESX, ESXi VMware Workstation or VMware player.  It should be installed, preferably on a third site, or at the minimum on another server/esx host/desktop PC completely seperate from the main P4000 environment.

The FOM integrates into a management group as a real manager and is intended only to provide quorum to the cluster, one of its main purposes is to provide quorum in Multi-site clusters.  The FOM does not need to be available for normal day to day operations, it is only used in the event of a quorum dispute between nodes.

Do I need one or not then?

In a two node P4000 cluster, you will need a FOM for automatic failover.  If no FOM is present then if one node loses connectivity, the other node also goes offline as well and you will have to manually failover the LUN’s.   If a FOM is in the same management group as the two nodes, the surviving node will know it is still on the network (as the FOM acts as an arbitrator node) and will stay online.

If you have an odd number of nodes, then you don’t actually need a FOM, as there should never be an even number of nodes available (unless you are really unlucky and have multiple node failures! which isn’t likely).

Feb 10

HP 3par… and?

So, your favourite(!) storage vendor has just shelled out 2.1 billion dollars on a product line that sits alongside their bestselling Mid-Enterprise Storage Arrays (EVA) but also pokes its finger at the Mission Critical arrays (XP).  Just what are customers going to gain from 3par?

3par offers:

  1. FC & iSCSI connectivity with built in dedicated site to site replication ports
  2. Storage Vitalization
  3. Thin Provisioning built in (its the first platform to offer this technology built into it)
  4. Thin conversion on the fly (and in the background)
  5. Has the capacity to migrate data between FC,SSD & SATA drives (Autonomic Tiered Storage)
  6. Multi-Mesh active-active controllers, which allows every LUN to be served by every controller
  7. Start small, grow in affordable increments
  8. Multi-Tenant system – run diverse workloads across the same platform
  9. High & Predictable Performance through 3par ASIC
  10. RAID5 & 6 without the usual performance impact, with the added bonus of Rapid RAID rebuild

A more than Enterprise level storage array that ticks all the boxes that many people need, especially in the cloud & utility market place.

Oh, and the ability to use words such as ‘chunklets’ & ‘raidlets’ in a sentence! 🙂