FF Focus

Once again, social media platforms are facing calls to tighten regulations on their platforms, following the hurl of racial abuse to members of the England football team. After losing to Italy in the final of the Euro 2020, certain players received swarms of abuse on social media, which critics say lies in the hands of these platforms to regulate.

London Mayer Sadiq Khan directly called on social media platforms to ‘act immediately to remove and prevent this hate’.

What kind of responsibility do these platforms have to prevent the spread of hate? Is there a way to leverage automation and machine learning to make this job easier?

Traditional media in the UK has an agreement with regulator Ofcom that makes them accountable for any form of abusive response to content. For example – a racist comment left on a BBC News article – the BBC would be accountable.

Ofcom doesn’t have this agreement with social media platforms because they aren’t considered to be publishers or broadcasters. These platforms remain self-regulating, so the question of accountability remains a grey area.

The difficulty these tech giants face is the huge volume of user-generated content, which swamps the efforts of human moderators employed by these platforms. It’s not expected that human moderators can sift through every piece of content, as soon as it’s posted, to see if it contains hate. The solution has to be automation.

We can already see social media platforms utilising automation to prevent misinformation around COVID-19. For example, Instagram immediately fags any content that contains information about COVID, and points users to the World Health Organisation for accurate advice. Critics have suggested this same technology be used to detect racist and abuse content.

Automation is already built into the algorithm in this way, such as a blanket banning of certain hashtags and words. But it can only do so much. Automation is currently unable to understand context, nuance, different cultures, etc. A certain emoji may not be offensive if said to one person, but when the context is different, the intention also changes. Instances such as this is where automation fails.

So what is the solution? Stopping trolls from posting hate on these platforms is of course, the ideal solution. But alas, an impossible ask. The more likely solution will take time – develop automation technology that is intelligent enough to detect the context of hate. Until then, there remains some power in the hands of users to report hate content when they see it. 

written by Evangeline Hunt

Several major websites went down on Tuesday morning, following a software bug on cloud-computing company Fastly.

Fastly said the bug was triggered by a customer configuration change.

The outage lasted 49 minutes, affecting some popular websites including Amazon, the Guardian, Reddit and even the UK gov website.

The problem originated with the cloud – more specifically the content delivery network (CDN) operated by Fastly. The Fastly CDN is a global network of servers, used by organisations to deliver content as quickly as possible (oh the irony).

Fastly were quick to recognise, apologise and resolve the issue. ‘This outage was broad and severe’, said Fastly’s Senior Vice President of Engineering and Infrastructure, Nick Rockwell. ‘We’re truly sorry for the impact to our customers and everyone who relies on them.’

Although the global outage was dealt with quickly, it does highlight how dependent many organisations are on cloud services and service providers.  

This didn’t turn out to be a cyberattack this time. But it does raise the question as to what would happen to any of these providers if they fell victim to a cyberattack. The consequences would be far worse than a 49-minute outage.

As we approach the halfway point of 2021, it’s a good time to reflect on the trends that were predicted for the year. Most of us were relieved to say goodbye to 2020. A year full of uncertainty and change. Not all of them were bad – the past year pushed engineers into a new way of collaboration to build, deliver and manage IT Infrastructure.

This digital transformation has become crucial for business success. As a result of the challenges of the past year the following DevOps trends have gained more attention:

1. In 2021 Service Mesh has increased and become one of the key components of the dedicated infrastructure layer built into an app as that’s how parts of an application can share data with another. Service mesh is used to facilitate service-to-service communications between services. Due to factors from choosing particular tools, people will be forced to use them. This means, as tools become a more and more inseparable part of other solutions, more service mesh will be used. On top of that, it will provide the features and standards within each application and therefore represent the platform for ALL kinds of applications.

2. DevSecOps is gaining more importance, ensuring security for businesses of any sizes as working with cloud-based technology is becoming part of our daily workspace. Vulnerabilities and security gaps need to quickly be noticed, detected and diminished by the DevSecOps team.

3. Kubernetes is an open-source container-orchestration system for automating computer application deployment, scaling, and management. It was originally designed by Google and is now maintained by the Cloud Native Computing Foundation. The use of Kubernetes continues to grow in 2021 as they are needed to build complex cloud-native infrastructures for automating computer applications. As these are difficult to understand, in 2021, cloud operators and practitioners create new supporting tools and will benefit the tech community. These tools focus especially on data science, visibility and secret management.

4. AI & ML driven DevOps approach: nowadays, traditional organisations need to handle a massive amount of data that is being generated with immense speed, variety and volume. To be able to analyze and compute this data of any scale and size, organisations work with AI (Artificial Intelligence) and ML (machine learning) as they are the boosters to transform the workflow of teams. ML helps to understand where blockages or capacity issues that occur in the delivery lifecycle and therefore improves developing, deploying, delivering and managing applications properly.

5. Observability or monitoring is the question. As systems are transforming into more complex, cloud-native and open source microservices running on Kubernetes, more engineers will be mindful of observing and monitoring the application to identify and respond to outages and events. This deeper insight into downtime is an approach to monitor, analyse, trace events to investigate the specific causes and pinpoint the impact on the company. 

Excited for the second half of 2021? Well, if you weren’t yet, then you should be now. The fact that DevOps is the best solution to increase quality, save time and money for companies hasn’t changed. But still, a new era is well underway in this cloud-centric and all-digital world. Therefore, upgrading and integrating techniques and tools are required for the rapidly changing market needs.

written by Sophie Finsterer

Businesses around the world are waving goodbye to datacentres in favour of the cloud, in order to innovate and grow to meet demand. As an IT infrastructure professional, having experience/certifications within cloud is going to make you more employable as cloud computing continues to rise in popularity.

AWS, Microsoft Azure and GCP are the three biggest cloud platforms used globally. IT professionals looking to upskill themselves should consider training in one of these three platforms – but which one will be the most valuable? We ran a poll on LinkedIn to ask that question – the results are as follows.

What is your preferred public cloud platform?

AWS = 55%
Azure = 38%
GCP = 8%

(218 people surveyed)

It’s no surprise that AWS gained the most votes – it is the market leader and the oldest established cloud service. According to Canalys, AWS owns 31% of the market as of July 2020 – with Azure at 20% and GCP at 6%. This is to be expected considering the seven-year head start that AWS had without any real competition.

But AWS isn’t just the oldest cloud provider, it also has the most services to offer. Its enterprise-friendly features make AWS a solid option for large organisations – such as Netflix, AirBnB, Nike and the Royal Opera House.

Microsoft’s Azure is closing the gap towards the market leader as it builds up its platform. It gets a lot of business from tech companies that already have a relationship with Microsoft, or who already use their programs such as Office365 or Teams. Microsoft can make it easy for these companies to transition to the cloud seamlessly, which is an attractive feature.

While Azure initially struggled to work with open source technologies, this has recently changed with around half of its workloads running on Linux.

GCP is the “new kid on the block”, which explains why it came in third in our poll. Yet it appeals to certain companies due to its strengths in big data, machine learning projects, and cloud-native applications. Despite this, Google has more work to do if it wants to compete with the likes of AWS and Azure.

So which cloud platform should you consider upskilling yourself in as an infrastructure professional?

‘AWS is still the market leader and the most popular, but Azure is catching up and so many businesses partner with Microsoft that it will make you really employable if you have skills in Azure’, says our Cardiff Consultant and cloud expert Gareth Streefland. ‘Plus, it’s probably the most approachable for engineers starting out with cloud.’

Upskilling yourself on any platform will improve your employability and job prospects. If you are experienced in cloud computing and are looking for new opportunities, feel free to get in touch – we would be happy to help.

Can you see a shift occurring in the preferred public cloud platform as time goes on? Or will AWS remain the market leader for the foreseeable? We would love to hear your thoughts.

 

written by Evangeline Hunt

When starting out in IT Recruitment you are confronted with numerous job titles that sound very similar at first but have distinct differences when you take a closer look. So, I found myself asking the question: “What’s the difference between a System Administrator and a System Engineer?”

My first understanding was the following: Being a System Administrator involves installation and configuration on already existing systems. System Admins take care of IT Infrastructure which were developed and implemented by a System Engineer. The infrastructure can consist of Servers, Networks, Desktops or all of the above.

After talking to both System Administrators and System Engineers, I then realised that the IT infrastructure world isn’t simply black and white. I spoke to a System Administrators who only focussed on the implementation of systems, others were taking care of automated Server landscapes, some were actively involved in the automation processes and another group managed the full life cycle, including concept, design and implementation of systems. On the other side I spoke to a similar variety of System Engineers.

Needless to say, I was confused. Were the two job titles simply a personal preference? Or a company choice? I didn’t understand the fine differences between the two.

Not willing to give up until I found an explanation, I searched many websites and blogs to find the answer. During the search I noticed that the topic was widely and heavily discussed, I found supporting evidence in blogs and discussions that the difference, in most cases, was simply the title. However, my feeling is that the difference between these two roles has narrowed due to the developing IT landscape. Nowadays, companies need hybrid engineers who can take care of the infrastructure maintenance work as well as the implementation and automation.

A case study from my own experience as a recruiter supports that opinion: I was working with two different candidates, one of them was a System Administrator and the other one was a System Engineer, however, they both had very similar skill sets including automation tools like Puppet and Ansible. The company I introduced them to were looking for a System Engineer. In the end the System Administrator was chosen for the job!

After all this time working with both System Administrators and System Engineers, I still haven’t truly found a distinct difference between them. Maybe our network can help me find the answer. Whether you share my opinion or disagree, I would like to hear your thoughts on this.

Cloud computing is massively on the rise in the current day and age. In fact, 81% of companies with 1,000 employees or more have a multi-platform strategy. 

Cloud technology has redefined the way in which companies store and share information. It has transcended the limitations of using physical devices. 

Cloud Technologies provides many benefits such as better scalability, better storage options, better collaboration with remote users and highly affordable for a lot of companies. 

But what does the future of cloud technology look like? 

Matt Riley CEO & Co-Founder of Swiftype commented “A decade from now, every business will be operating primarily from the cloud, making way for more flexible — yet more productive and efficient — ways of working. Hardware won’t be the problem in a decade — software will.” 

The future is bright for cloud computing. Analysts at IDC estimate that the field will evolve rapidly in the coming years, with almost 75% of data operations carried out outside the normal data centre. Moreover, 40% of organizations will deploy cloud technology, with edge computing becoming an integral part of the technological setup. Also, a quarter of end-point devices will be ready to execute AI algorithms by the year 2022. 

Cloud Computing trends on the rise - automation 

The automation tools available to us have proved to be very important when it comes to addressing errors in business processes, meanwhile streamlining them to generate fruitful results. 

For instance, developers can make changes to their websites hosted on the cloud before going live. If anything goes wrong, they can restore an older version of the website without affecting the sales process or user experience. As soon as the website goes live, it starts getting traffic. 

Opting for cloud means there will be more data consumption involved. Managing applications and routine tasks can become tedious. Developers can use automation to get rid of the manual process they have to use to carry out daily operations. 

Serverless paradigm 

The serverless paradigm is the next revolution in waiting, according to the CTO of Amazon. The concept of serverless paradigm relates to the fact that it facilitates cloud to execute a code snippet without any hassles for the developers. 

Using this approach developers can divide software into chunks of code to upload on cloud to address customers’ desires, thereby delivering valuable experience. This practice ensures faster release cycle for software. Amazon Web Services (AWS) has already started using the serverless paradigm to its advantage. 

----------

As cloud computing continues to make inroads in enterprise worlds, all stakeholders are looking forward to the evolution of the model. As things stand today, almost every significant innovation such as blockchain, artificial intelligence, AR/VR, robotics, and IoT rely on cloud computing technology. 

It’s not just computational power, networking speed, or storage capacity that makes cloud computing great. Those are just operational metrics that better technology would eventually change and replace over time. The real value of technology is what it does, not what it’s made of.