Posted by Chris Evans on 12 November, 2009
Here are a few pictures from tonight’s dinner; some people turned up later than planned and joined in as they arrived. Congratulations to the restaurant staff who served them so quickly. Also congratulations to those who won prizes from Stephen’s cryptic competition. We’re expecting good things from Devang as he won again with video hardware. Let’s see some videoing!
Posted in Uncategorized | Leave a Comment »
Posted by Chris Evans on 12 November, 2009
Finally arrived at the DoubleTree in San Jose. Met up with Devang, Nigel and Simon at the airport and shared a leisurely hour discussing our thoughts as we travelled to the hotel. Thanks to Stephen and Claire for a great limo!
Whilst checking in we bumped into Robin Harris in the foyer. I’ve read Robin’s blog for quite a while and it was great to finally meet him in person.
Now it’s time for ablutions and off to the bar in a vain attempt to keep awake (8 hour lag for us Brits). More later, possibly including some photos from the bar (if they’re printable!).
Posted in GestaltIT | Tagged: Devang Panchigar, Gestalt IT, Nigel Poulton, Robin Harris, San Jose, SImon Seagrave, Tech Field Day | Leave a Comment »
Posted by Chris Evans on 11 November, 2009
The day has finally arrived and I’m sitting in the BA Lounge in Terminal 5. As with any long distance trip I have that usual mixture of trepidation and excitement, especially when travelling to a place I haven’t visited before.
T5 is Heathrow’s new flagship terminal and the dedicated home for British Airways. Unfortunately, my experience to date travelling to and from T5 has been less than spectacular. So far, today things seem to have been a little easier.
So, looking forward to the remainder of today, what can I expect to be in store for me and the other Gestalt IT attendees?
Well, after landing and heading off to the hotel, we’ve dinner planned for 7pm with a chance to catch up with everyone. Of course for us UK folk, that will already be 3am UK time. We then have to leave the hotel by 7am the next morning – yes 7am to go to the VMware offices. It’s pretty clear that we’re going to have a busy time.
I’ve compiled a list of some of the things I’m hoping to achieve or learn about whilst I’m here. I know some of the vendors; others I have no knowledge of at all. Here’s what I’m looking for:
3Par – information on the new “Thin built-in” technologies, which were just supported in software; real world calculations on TP savings; array clustering; solid state and block-level tiering; VMware integration; vStorage API support; CLI/API interfaces.
Data Robotics – more information on proprietary RAID; management framework integration; hot swap operation (graceful rather than forced); futures (larger devices, rack mount, FC(oE) support); DroboShare futures (AD integration, iSCSI support)
Nirvanix – CloudNAS; security; block and file protocol support.
Ocarina – Value proposition (cutoff where de-dupe savings recoup hardware investment); platform integration (e.g. BlueArc); compression ratios; data replication between archives; imgestion rates; ingestion methods.
Xsigo – market positioning – fabric or array; perceived competition; TCO/value proposition.
Whatever happens, I’m sure it will be a fascinating trip. Look out for more updates here and on Twitter (@chrismevans)!
Posted in Uncategorized | Leave a Comment »
Posted by Chris Evans on 6 November, 2009
The GestaltIT Tech Field Day is only few days away now and you can expect by this time next week we’ll have some interesting posts on the seven presenting sponsors. By way of a warm-up, we’re providing a competition, full details of which can be found here. So, follow the links to one or more of the specific companies below and fill in the details. You never know, you could win…
Posted in Enterprise Computing | Tagged: 3par, GestaltIT, MDS Micro, Symantec, Xsigo | Leave a Comment »
Posted by Chris Evans on 4 November, 2009
Barely a day has gone by since EMC, VMware and Cisco announced their partnership, and now we have a view on HP’s strategy. Today, HP announce their private cloud offering – Converged Infrastructure Architecture (CIA), putting them squarely head-to-head with the VCE coalition. So what’s it all about?
HP already offer server, storage and networking infrastructure. The CIA (nice acronym) brings those technologies together and leverages some of their recent acquisitions. There are four main “innovations”:
Looking specifically at the Virtual Resource Pools, there are three storage-related offerings:
StorageWorks X9000 Network Storage Systems Family. This is the technology acquired from IBRIX and is being sold in three flavors; X9300 gateway – essentially a NAS head; the X9320 appliance, using SAS/SATA drives; X9720, the high-scaling model. What’s nice here is that any and all of these offerings can all exist under the same namespace, with options to rebalance and move data dynamically between nodes.
StorageWorks SAN Virtualisation Services Platform. I have to admit I know little nothing about this product, however it looks like an IBM SAN Volume Controller equivalent, pooling and representing fibre channel storage resources. I first saw mention of it on Stephen Foskett’s blog about the HP Tech Day (had I been able to make the event, I’ve heard about it in person).
StorageWorks Cluster Extension EVA Software For MS Hyper-V Live Migration. This feature provides support for Live Migration – the equivalent of VMware’s VMotion.
Opinion
HP is clearly setting out their position to compete head on with the VCE coalition. What’s interesting is that this first announcement discusses new storage technologies, rather than the existing XP and EVA ranges. I’d expect these are not going away, but it would be good to understand how they fit in the overall strategy. Even more telling is the lack of comment regarding integration at the hypervisor level. Does this mean HP will offer both VMware and Hyper-V? Note that the first O/S announcement related to Hyper-V support…
IT is moving to the converged model. We are seeing a ‘”land grab” for the customers hearts and minds and for the large players to entrench their solutions in the data centres of the future. What’s not clear is where some of the other big players sit. IBM can do stuff alone; but what about HDS and Netapp? At the moment they seem left out in the cold.
I look forward to finding out more on these products in the coming months as all the vendor strategies start to pan out.
Posted in Enterprise Computing | Tagged: Converged Infrastructure Architecture, HP, IBRIX, Stephen Foskett, SVSP, X9000 | 3 Comments »
Posted by Chris Evans on 3 November, 2009
The time has come to get a “proper” design for my blog site, now that I’m firmly established with WordPress. Currently I’m using the standard WordPress “Cloud” offering on WordPress.com and will look to enable and pay for custom CSS access.
So, I need some help on creating a new look and feel, which will meet the following requirements;
If you’re a WordPress whiz, then take a chance and do me a design. I have 500 english pounds (good currency these days) on offer, payable via Paypal.
Drop me an email or Twitter DM if you’re interested.
Posted in Uncategorized | 2 Comments »
Posted by Chris Evans on 3 November, 2009
Finally all the speculation is put to rest as VMware, Cisco and EMC announce their joint venture: Acadia. The “coalition” of the three companies will work together to deliver private cloud infrastructure which utilises Cisco UCS, VMware vSphere and EMC storage hardware. All of this will be delivered in something called the Virtual Computing Environment (VCE and conveniently the initials of the three companies) and in packages called Vblock (interesting that the ‘V’ is capitalised here). Acadia acts as the delivery company for the technology and from the legalese on their website seems to be an EMC entity.
I can’t help thinking back to my mainframe days and the monopoly held by IBM in the development of the entire System 360/370 & 390 architectures. Many companies copied (Amdahl, HDS) and EMC effectively broke the grip IBM had on storage hardware. Now the mainframe is niche and things have moved on, however EMC seem to be heading full circle by packaging their technology in the triumvirate that is VCE, but is perhaps more a duumvirate than we care to think. How ironic.
So what is the initial offering? Well, there are three offerings labelled Vblock2, Vblock1 and Vblock0, covering the support of up to 6000, 3000 and 800 virtual machines respectively. High end uses V-Max storage, the middle offering uses CLARiiON and the low-end uses “EMC Unified Storage”, which presumably is Celerra. Ionix rears its ugly head as the management tool of choice, packaged as “Ionix Unified Infrastructure Manager”.
Whilst some customers will find the packaged delivery of the components they may well have already chosen as helpful, what’s not clear is how readily other technologies can be integrated into the Vblock architecture. For instance, if EMC storage isn’t the preferred option or if Hyper-V is the hypervisor of choice (regardless of how unlikely this may be), how can customers implement their configuration? There’s no incentive for VCE to offer other options and that just reduces customer choice.
How will the existing VMware partners like HP and Netapp feel? No doubt we’ll see other partnerships and offerings spring up and I don’t think that will take too long, however some vendors will quickly be left behind.
Ultimately, we all know that IT is moving into a new phase where virtualisation of all components is the key strategy. Networking, servers and storage are all moving to be commodity and consequently there needs to be another approach to making money and achieving high margin from hardware and software sales. Delivering a consolidated architecture where integration means the sum is greater than the parts is what ACE and Acadia will deliver.
You can read the Cisco press release here. Scroll down to the “Additional Resources” section for more links.
Posted in Cloud computing, Enterprise Computing, GestaltIT | Tagged: Acadia, Cisco, EMC, VMware | 2 Comments »
Posted by Chris Evans on 28 October, 2009
This is the second of a series of posts reviewing the StarWind Enterprise Server iSCSI software target. The posts relate to version 4 of the product. Previous post(s) can be found at:
Review: StarWind Enterprise Server – Part I
In the first post of this review, I explained how StarWind’s server product provided the ability to turn a Windows server into an iSCSI storage device. In this post, I’ll focus more on the specific LUN types available to the user.
There’s not doubt the range of supported devices is comprehensive, although some features such as encryption and compression need more explanation to understand how they are implemented and how data could be recovered in a server failure scenario.
Thin Provisioning
It’s pretty much taken for granted that thin provisioning needs to be provided as a feature in storage arrays, whether physical or virtual. Fortunately StarWind Server offers this as part of the CDP device type. As a test, I created an iSCSI thin LUN using the console and connected it to a Windows host. As expected, copying files to the LUN grew the device by the size of the copied files. Unfortunately, deleting the files and/or reformatting the drive did not reclaim the released space. In fact, a full format of the drive, rather than a quick format resulted in the full capacity of the drive being allocated and the benefits of thin provisioning wasted. (Tip: Use Quick format on Windows). In this release of the product there appears to be no way to reclaim unused resources within thin provisioned devices. Thin Provisioning implementations need to add this functionality as the value of using thin LUNs simply diminishes over time.
Summary
Overall, the product offers a wide range of device types. A more detailed understanding of some specifics (e.g. encryption standards) is required. Thin LUNs need some space release technology added to ensure thin doesn’t simply revert to thick over time.
In the next post, I’ll look at security, performance and the competition.
Disclaimer: I have no connection or relationship to StarWind Software. A free copy of the StarWind Server software was provided by the company for this evaluation.
Posted in Enterprise Computing | Tagged: iSCSI, StarWind, StarWind Enterprise Server, thin provisioning | 2 Comments »
Posted by Chris Evans on 26 October, 2009
This is the first of a number of posts reviewing the StarWind Enterprise Server iSCSI software target. The posts relate to version 4 of the product.
Not everyone wants or can afford to buy dedicated storage arrays, especially in current climates. So if you don’t want to spend money on a bespoke storage solution, are there any other options available? One of them might suit your budget – StarWind Enterprise Server from StarWind Software.
Background
Classic SCSI predicates a client/server architecture for storage transactions, where the client or host is known as an initiator and the storage device as the target. This is replicated in the iSCSI architecture and today we already see free iSCSI initiator implementations in host operating systems such as the Windows family.
StarWind Enterprise Server is a software product which turns a traditional Windows server into an iSCSI Target, meaning the server can act as a storage device, serving out LUNs to other servers across the IP network. These LUNs operate exactly as iSCSI devices would if they were presented from a more expensive and dedicated storage array.
The Technology
StarWind Enterprise Server installs two components; the Windows service and the management console for creating and administering iSCSI devices. Administration is pretty simple; see the screenshots below. One shows my list of devices and the second shows the starting process to creating a new iSCSI LUN, of which there are currently eight types:
When any of the above devices are created (excluding the RAM device), the new device is mapped to a number of files on disk. An Image File Device is created as a simple flat file with the extension .img. A Snapshot and CDP Device is created with three separate files to hold the structure, as these devices can also act as thin provisioned LUNs. I’ll discuss more on the various flavours of LUN type in a future post.
Now the first question has to be why bother using this software? In large organisations with dedicated equipment then that question is a fair one, however not every company wants or can afford to deploy storage arrays across their environment. There are scenarios where block-based storage is still required, however cost of acquisition and management precludes the deployment of traditional solutions. More important, the StarWind iSCSI target can be deployed as part of a virtual environment, leveraging the storage deployed in say, a branch VMware server to provide centralised storage resources.
So, this first post is to set the scene. In coming posts I’ll be looking in more detail at LUN allocation types, the thin provisioning aspects of StarWind Server, security, the competition and of course performance.
Disclaimer: I have no connection or relationship to StarWind Software. A free copy of the StarWind Server software was provided by the company for evaluation.
Posted in Enterprise Computing | Tagged: iSCSI Target, StarWind, StarWind Enterprise Server, thin provisioning | 3 Comments »
