As I write this, it has been just over four weeks since Enix acquired Brixly. A fair amount has happened since then, and as has been correctly pointed out to me, communication is key. So, here is a bit more of an introduction to Enix, our plans, and what has transpired in the last four weeks.

Some of the following points reiterate my comments on the feedback board. If you have read those comments, some of this may sound familiar, but I think it’s best to include them here for everyone to see.

A Brief Enix History

Enix was started in 2006 by myself and my business partner Nick. We began as PHP developers working on mission-critical sites with extremely high traffic, gaining substantial experience in high-availability application development and infrastructure management.

Dream Hosting started very small at first under the domain dream-hosting.co.uk. We purchased this domain and a handful of hosting clients from someone we used to work with, initially as a side hustle. Over the years, it has grown significantly. We’ve made plenty of mistakes and learned a lot along the way.

Around 2013, Dream Hosting (and several other domains) combined into hostpresto.com, which has become our primary brand. Since then, we’ve gradually made acquisitions, either keeping them as standalone brands or merging them into hostpresto.com. As of now, we have four primary hosting brands (have a look on enixltd.com).

Staff

As we’ve grown, our team has expanded as well, with each acquisition bringing in more extremely talented staff. Here is a brief introduction:

• Elliot - CEO/CTO: I push things forward, manage the business side, and head up the infrastructure. I am an experienced sysadmin and developer.

• Nick - COO: Ensures smooth operations and manages day-to-day activities. He is a seasoned developer and sysadmin.

• Dennis - CXO (From Oct 2024): When Dennis returns, he’ll resume his role in communications, support management, and strategy without the additional headache of running the business, ensuring you get the best possible experience.

• Mike / Sam / Liam / Danny: 2nd Line and Managed Services Support. Experts in web hosting and server management of all types.

• Declan, Chris, Andy, Rob, Harry, Paul, Wesley, Tom, Shabs, Faizal: 1st Line Support. All of them are extremely experienced, hardworking, and knowledgeable.

• Development and Engineering Team: Behind the scenes, Enix has a team of ten developers and engineers working on new services, products, optimisations, and maintaining operations.
  

The Recent, Controversial Email Changes and Reasons Why

The decision to remove MailChannels and SpamExperts was made early in our plans for Brixly, following extensive discussions with Dennis. Brixly provides a premium service at a non-premium price, and Dennis has done a remarkable job getting it to this point.

I won’t hide the fact that SpamExperts and MailChannels are very expensive, and it’s a cost that needed to be removed. In the interest of transparency, Brixly needed help, and if it wasn’t Enix, it would have been someone else. Brixly is a premium host at a non-premium price. Customers get a lot of value, but Brixly cannot survive at its current cost structure.

Losing Brixly would be a great shame, and our acquisition is not a cash grab; we’re trying to get it into a sustainable situation to continue growing and improving it. I truly believe Enix was the best choice for everyone involved, and Dennis agrees. We are both incredibly similarly minded regarding web hosting, aiming to achieve the same goals in the same ways.

Email hosting in the shared hosting and reseller environment is challenging. How can you host mission-critical email at an average cost of £1/month per website (or websites if you allow addon domains) with virtually unlimited mailboxes on a reseller package? It’s not easy, considering Office 365 and Google Workspace cost £6/month+ per email account. We will always do our best to provide top-notch service, but managing expectations is essential.

While the decision to remove MailChannels and SpamExperts may seem drastic, it was necessary. It will be the most impactful change on customers, and the worst of it is now behind us.

At Enix, we’ve been running cPanel and related infrastructure on a large scale for some time. We’ve used a heavily tweaked version of MailScanner for around ten years. Additionally, we have comprehensive systems to monitor and filter outgoing email, taking action on spam. Customer issues regarding incoming spam or email delivery are minimal, if any. It’s a great system that works well for all concerned.

Is MailChannels and SpamExperts better than our new solution? That’s subjective. We’ve had a mix of customers: some desperate for those services to return, and others preferring the control our solution provides.

MailChannels had a couple of major outages recently, making us uncomfortable due to the lack of control. What if the outage lasted a couple of days? A week? There’s also the issue of data security. Using third parties for email exposes us to risks. If Enix is compromised, that’s on us. If a third party is compromised, it’s still on us.

Confident in our ability to provide reliable email without third parties, we decided to remove MailChannels and SpamExperts. This aligns with our Enix ethos: to provide great hosting at the best prices and to do this for the long term by owning the technology stack fully.

Key Points of Our Ethos:

1. Control: We avoid relying on third parties for services where possible, managing them ourselves to avoid dependency.

2. Future-proofing: By building and maintaining our systems, we have full control and can implement improvements with long-term predictability.

3. Financial Stability: The cPanel buyout led to a significant cost increase for hosts worldwide, coupled with increased energy costs and inflation. Relying on third parties like MailChannels or SpamExperts means we cannot control those costs.

This shapes our long-term plan, which I’ll highlight further on.

What Went Wrong?

Despite thorough planning, we didn’t foresee Microsoft blocking our IP ranges immediately. We barely sent an email before they did this. We’ve opened around 200 tickets with MS to have the blocks removed. Despite our clean servers, they didn’t like the sudden increase in volume from our ranges.

Over the past week, we’ve had increasing success with deliveries. Our anti-spam measures keep our servers at less than a 0.1% complaint ratio, the lowest MS report.

During this time, support tickets overwhelmed our team. I brought in emergency extra support, but it was sometimes a losing battle. I have had very little sleep this past week, and the team has worked tirelessly. While our support quality has slipped, I’m proud of what our team managed given the circumstances.

The issues were compounded with incoming spam performance. It takes time for the spam system to warm up and learn. In hindsight, we shouldn’t have made both incoming and outgoing changes simultaneously, but we’re learning from this.

The issue is improving hour by hour, and we’ll work tirelessly until it’s fully resolved. I apologise to all customers affected by the recent issues. Please know that we have taken it seriously and done everything within our capability to resolve the issues as quickly as possible. I thank everyone for their understanding and patience.

This is not going to be a regularly occurring event. This was the single biggest and most impactful change we wanted to make. Normal service will resume. We have a couple more changes coming, but they will be relatively minor, which brings me to the next point.

What the Hell Are You Going to Do to My Service Next?!

One more significant change is coming, but it’s not on the same scale as the email change.

Brixly currently rents all hardware from various suppliers across the world. Enix, however, owns and operates all our hardware for stability, control, and financial optimisation. Removing third-party costs is necessary for Brixly’s viability and allows us to invest in improvements.

Brixly is almost treading water, and we want it to move forward like a cruise ship—steady and sure, with plenty to offer and perhaps a little fun.

In 6-8 weeks, Enix will start migrating customers to our infrastructure from the rented infrastructure. This migration will be done slowly over the course of 6-12 months. These will be like-for-like migrations (cPanel or DirectAdmin) to servers with the same configuration and packages. Some customers have already been migrated to Enix infrastructure to sidestep the email issues, and new Brixly customers have been deployed to the Enix infrastructure for the last four weeks.

Both Enix and Brixly have significant experience with these migrations, and they will be effortless for you. The server hostnames will remain the same, but the IPs will change. Migrations will be scheduled, and you’ll be notified well in advance.

The bonus will be the new hardware. As part of the Brixly acquisition, Enix has committed over half a million pounds to new hardware for our next-generation infrastructure. The latest AMD Epyc CPUs and NVMe SSDs will be used. Here’s how it will compare to the current hardware:

• Newer AMD Epyc Milan CPUs

• Enterprise-grade, high-capacity NVMe SSDs forming a 1.5PB storage cluster

• High-speed, low-latency networking

• Ceph Storage cluster with triple redundancy vs the current 2x

• Backups to NVMe SSD storage for ultra-fast backups and restores (critical in a DR situation)

• Highly optimised for less power usage and powered by 100% green energy

For our customers in the US and EU regions, the same hardware will be deployed, offering an even greater improvement over the current rented hardware, and we aim to increase our presence in these regions over time with a larger product offering.

Long-Term Plans

Our team of ten developers and engineers has been working on projects behind the scenes for the last few years. Here’s a highlight of what’s coming, although you won’t see any of this until at least Q1 2025, once migrations are complete.

• Enmail: A new, purpose-built enterprise-grade email cluster for mission-critical email. We’ve taken the best in open source and built a new email cluster on top of Kubernetes for infinite scalability, MailChannels-like outbound load balancing, and more. This will be offered free to all customers to replace cPanel-based email.

• Encuro: A billing and support system designed to surpass all others. It features a comprehensive API for reselling all products, a comprehensive migration centre, whitelabel for resellers, and full integration into WHMCS.

• Envision: A cPanel-like hosting platform built on Kubernetes for ultimate performance and flexibility. It offers high-speed caching out of the box, optimised PHP processing, and complete scalability and redundancy for 100% uptime.

• Generally: We will improve customer experience and flows, including support, migrations, documentation, and all of the other ‘basics’.

We won’t force any services or products on anyone. We know many of our customers prefer cPanel or DirectAdmin, and these will always be options. We want these services to add value for you, not just for us.

In Conclusion

I started this update at 6 AM this morning, and now my fingers hurt. I should probably check in with the rest of the team. I hope this sheds some light on us as a company, our thought process, the reasons for our actions, and what’s coming.

As I mentioned on the feedback board, Nick, Dennis, and I truly want Enix and Brixly to be the best hosting service available, and we’re committed to keeping it there for the long term.

I’ll make every effort to update as regularly as required on changes in as much detail as possible. I hope for all of our sakes this will be it for a little while at least!

Elliot