vSphere 6.0- time to upgrade

vCenter-logoIf you’re running VMware vCenter and ESXi 6.0 it’s time to start planning to upgrade as General Support ends on 12 March 2020- one year from now and five years from it’s release. Thankfully the upgrade from 6.0 to 6.5 or 6.7 is usually quite straightforward, and VMware have put a lot of work into streamlining this process.

Looking at the Product Lifecycle Matrix other notable products in the VMware stable worth keeping an eye on include NSX for vSphere (NSXv) 6.2, Site Recovery Manager (SRM) 6.0 and 6.1, and vSAN 6.0-6.2.

vRetreat February 2019- Secondary Storage with Cohesity

Last week I had the pleasure of attending the latest #vRetreat blogger event. This edition featured a day of presentations and labs from enterprise storage vendor Cohesity held at Chelsea Football Club in London. In my first blog post from the event I look at what Cohesity are doing to distinguish “Secondary Storage” from “Backup Storage”.IMG_20190222_105103399_sm

There’s a number of vendors on the market who can provide enterprises with a backup appliance and support for public cloud storage. Cohesity have looked at this and asked what other business operations can leverage this (comparatively) cheap storage media? I’ve heard their message of “we’re not backup, but secondary storage” before, but at this event the distinction really clicked with me.

IMG_20190222_110747824_smWhilst front-line production services often demand the best-performing storage possible, storage for backups doesn’t (hopefully) need to be accessed regularly and doesn’t require the speed of access that front line production systems might. Where possible organisations will purchase cheap(er) storage for this task, and this can lead to a separate backup storage silo.

If nearly 80 percent of stored data goes unused after 90 days then the majority of data on NAS/SAN filers also fits these access and performance characteristics, so why not combine the two and reduce the silo count? The Cohesity platform offers SMB and NFS, and can also function as an object store. This also helps justify the outlay on the storage for backup which, like an insurance policy, you hope to never actually need.

CohesitySimilarly test and development workloads can often (but not always) be run on lower performance storage than their production counterparts. Again these functions are looking for similar attributes to backup when it comes to storage- keep the cost/GB low and don’t impact on the performance of our primary production storage.

Cohesity’s DataPlatform consolidates the traditional backup storage platform along with the ability to spin out test and dev workloads directly from this data, whilst also providing host file and object storage. For example, when the primary storage is upgraded to all flash, the NAS or test workloads that don’t need this level of performance can use the Cohesity platform.1550869582007_sm

This was an interesting briefing, and for me this part definitely showed the potential for not thinking of your backup infrastructure solely as an insurance policy but continuing to find new ways to leverage that investment elsewhere in the IT function.

Please read my standard Declaration/Disclaimer and before rushing out to buy anything bear in mind that this article is based on a sales discussion at a sponsored event rather than a POC or production installation. I wasn’t paid to write this article or offered any payment, although Cohesity did sponsor the lunch, T-shirt, and stadium tour at the event. Attendees were also given a pair of bright green socks and matching branded shoelaces so you should be able to spot them.

Powered Off VM cannot be Powered On

Symptoms

A powered off VM on ESXi 6.5 will not power on and returns the error “Failed to power on virtual machine…. The attempted operation cannot be performed in the current state (Powered off)”.

(i.e. the VM cannot be powered on BECAUSE it is not powered on!)

2019-02-19 (12)

Prior to being powered down the VM properties had just been modified. In this particular case it was immediately following a manual Ubuntu install and the install DVD (from a datastore ISO) was disconnected and the CD Drive switch to” “Host Device”. These operations were performed from the ESXi Web interface.

Repeated attempts to start the VM all fail the same way.

Solution

Unregister the VM, then locate the vmx file in the datastore and re-register it. The VM should now power on.

image

IT Blog Award Winner

Wow.

Thanks to all those that voted for “IT Should Just Work” in the IT Blog Awards and passed the devious maths test at the bottom of the voting form. I’m shocked and humbled to let you all know that the results are in and this blog won the award for Best Analysis. It’s surprised me, and the flurry of social media activity through the weekend following the announcement has shown there’s some big winners in the other categories.

Cisco Blog Awards Winner

I’d recommend anyone reading this visits the IT Blog Awards page and checks out the other winning blogs. Congratulations all!

To quote the list:

Thanks again- it’s always good to know that someone reads what I write and (hopefully) finds it useful or informative.

AWS re:Invent comes to London

2019-01-22_10-29-04Last week (January 2019) I attended an AWS event in London to learn a bit more about Amazon Web Services and catch up on the big announcements from the November re:Invent conference held in Las Vegas. This is my write up of that one day event held at the Altitude360 tower in the capital.

Keynote

After an (unusual) request for no phones or photography during the session the day kicked off with a Keynote from Chris Hayman. Chris started with the overview of AWS’s majority market share and highlighted some of their millions of customers. Pie charts and graphs showed just how dominant Amazon is in this sector.

He then went on to touch on the large portfolio of products in the AWS stable- this featured a slide with so many product names that even sat in the front row looking at this cinema sized screen I wasn’t able to make them all out. This gamut continued throughout the day- by lunchtime I estimated I’d been introduced to at least fifty different products from the range. This is all a bit overwhelming- “how am I going to learn about all of these?” I was thinking.

But then I had something of an epiphany- stop thinking of these as discrete apps like I would with a traditional server setup and instead think of them as components or features of the overall single package. For example-  in AWS you have a product called “Amazon Route 53” which provides a DNS service, there’s “AWS Transfer for SFTP”, “Amazon Elastic Block Store”, “AWS Systems Manager”. Back on your Windows Server you’d refer to these as the DNS, IIS, iSCSI Target, or Server Manager roles or services. This isn’t to say they are all equivalent between vendors, but thinking of these AWS product lines as services under the same umbrella platform helped bridge the gap from the more traditional way of thinking.

This proliferation of names and features does make writing up the conference without it becoming a list of products a little tricky so this post will focus on my highlights from the day with links to product pages etc. where appropriate. If you do want a full list, the “Products” link at the top of the AWS website will fill you in.

Along with a high level view of the announcements from re:Invent, the keynote also included a few customer stories- EMIS group who provide software and IT services to the NHS, the London Borough of Waltham Forest who are using AWS to leverage the data they collect to help citizens in their region, and LGSS Digital who are using Amazon Connect to provide AI powered call centres for local authorities.

The IT Crowd. Pic from channel4.com

Techies love the basement life

Following the Keynote the day split into a Business and Technical track. I favoured the latter, opting to stay in the windowless basement theatre when the Business track left for the 28-floor elevator ride to their views across London. Better to focus on the content rather than be distracted by the busy cityscape below 🙂

Core Services

The first session covered the Core Services in a bit more detail, starting off with the EC2 compute power that everything hinges on. We delved into the instance types available, from the new Graviton ARM processors (A1) through to the high-end compute with 4GHz Xeon cores (z1d).

In the storage part of the talk there were a couple of new features that caught my attention- notably the Managed File System offerings- FSx. The Windows version of this provides an AD-integrated DFS file-share with all the multi-availability-zone global resilience that the public cloud offers. A second service of note was the Glacier Deep Archive which offers storage at about $1/TB/month with a 12 hour recovery time for all those files that have to be kept for long periods but won’t be needed in a hurry.

Networking and Security came up next- one of the interesting things here was the “AWS Security Hub” which brings together all the AWS security products in one place. There’s a growing number of discrete products in the AWS security sphere (as with their other product groups) so having one system to manage them all makes a lot of sense.

Dev Tools and Containers

The afternoon sessions started with one covering developer tools, containers, and microservices. A big takeaway for me here was how AWS supports a development environment where multiple teams are using multiple languages to develop these microservices that need to all talk to each other for the application as a whole to work. The complicated nature of this needs management and visibility of the components as well as consistent communications between them. Features such as the preview “AWS App Mesh” and “AWS Cloud Map” can help developers with this challenge.

DB, AI, and ML

A quick dive into the world of databases, data lakes, data warehouses, ledgers, and blockchain came next. Again, every variant comes with a product name and there’s some inconsistency here- some products such as “Amazon Managed Blockchain” or “Amazon DocumentDB” have a name that clearly shows what they do, whilst others such as “Amazon Neptune” or “Amazon Aurora” leave you guessing. Thankfully they haven’t all been reduced to acronyms across the marketing and documentation -we have S3, RDS, EC2 etc but fifty+ of those would definitely get confusing.

My highlight of the section on AI and Machine Learning was our quick look at the the “AWS DeepRacer”- a scale model car which you can go out and buy and then teach to drive itself. During the break we got to see it whizz up and down the stage.

 

View this post on Instagram

 

An #Amazon #DeepRacer being put through its paces.

A post shared by Chris Bradshaw (@startmenu) on

Internet of Everything Else

We were next introduced to the wide range of IoT services available on AWS. These are split into 3 categories: Data, Control, and Device. There’s lots of interesting developments happening here with Amazon working with the endpoints collecting the data- perhaps integrating with legacy equipment using “AWS IoT Sitewise”- and then storing, before analysing and acting upon it.

The final session was left to cover “Everything Else”- all those things that didn’t fit into one of the above categories. This included the VMware supported AWS Outposts where Amazon hardware can be hosted in your own datacentre and, bizarrely given the audience, a look at the AWS “Groundstation as a service” offering in case you have a satellite you need to talk to.

Summary

Overall this was an interesting excursion into the world of AWS even if it was solid PowerPoint for every session. The opportunity to drink direct from the firehose of information and then have face to face discussions with the team from Amazon on how to apply it in my own environment was definitely worth the time, and I look forward to finding out more. Details on their future events can be found here: https://aws.amazon.com/events/