LowEndBox - Cheap VPS, Hosting and Dedicated Server Deals

Interview: Q&A with Ditlev Bredahl, CEO of OnApp

Tags: , , , , , , , Date/Time: September 2, 2020 @ 12:00 am, by raindog308

Our LowEndBox Interview Series has featured many interesting personalities so far and today we’re excited to bring you an interview with the CEO of OnApp, Ditlev Bredahl.  The name OnApp will be familiar to nearly anyone who’s ever worked in the hosting industry.  You may recognize some other famous names in Ditlev’s past, such as vps.net and SolusVM.  Ditlev has been involved with computers since the days of the Timex Sinclair and you’ll find his perspective on the hosting industry and where it’s going to be both well-informed and fascinating.  Let’s go!

Where in the world do you live?

I live in London, UK – have been here for many years now. Originally I am from Denmark, but I have worked or lived in countries around the world.

Tell me about your background and your company’s history

I’ve always been involved in tech – being 47yo my very early IT experiences started with ZX81, Spectrums, Vic20 and my beloved C64 (which I still have somewhere in storage). I’ve been programming or building hardware all my life – but never really been super good at either. My strength seems to have been in the grey area between business and tech. Though, I am happiest building new products and coming up with new ideas. 

OnApp was a spinout of UK2 Group. At UK2 we built VPS.net, and the core of OnApp was what we used to manage that platform. It was the golden days of cloud, I remember Carlos (Rego, now CVO at OnApp) and me getting VPS.net from 0 to 1m USD run rate in about 5 weeks. We had long waiting lists of people that wanted to buy cloud, at some stage we even asked them to write a poem about VPS.net to get to the top of the list…we got 100s of poems :)

At UK2 we also started the high-bandwidth server industry (possibly together with FDC), some funny threads on WHT here and here when we built the 100tb.com brand.

After selling UK2 Group to Lloyds Development Capital (LDC.co.uk) we transformed the VPS.net software stack into OnApp. AWS was starting to become a real factor in our industry, and we were on a mission to provide the industry with the tools that could enable them to fight back. A good strong platform at a good price with great support. That was OnApp then … and hopefully OnApp today :)

The first years of OnApp were pretty hectic…it was all about growth, hiring talented people and getting new releases out. Fast forward to today, we are now kind of a grown up company, with investors, board meetings and +120 employees across our 3 locations. We have a very strict QA structure, and release schedules…we can still release new products and bleeding edge features, but with all the large customers relying on our platform now, everything is just a little harder. 

I get it, we really can only release super solid code today – but I do miss the old days :)

Let me see if I understand your market position. The “big boy clouds” like AWS, Azure, and Google have fleets of PhDs to write their own cloud hosting software. Your software is designed for the next tier down, to allow both public hosting providers and private enterprises to build (and possible sell) cloud services. Am I close? How do you see OnApp’s market position?

Well, yes and no – first of all, +95% of our revenue comes from service providers. We eat sleep and drink hosting and, like me, most of my teams have a background in the industry and we are still on a mission to keep the scene vibrant and alive. Yes, we do have a few enterprises that buy direct from us, but about a year ago we spun out that arm of our business (check out sunlight.io) – I felt we had lost focus on our service provider customers, so after moving that part of our business to Sunlight we are now back with a laser sharp focus. I started my first hosting company in 1997 – this is the industry that made us, and we stand behind it 100%. 

True, AWS, Azure and Google will probably never use our software, that being said there are many of our service provider partners and customers that use our software to spin up and manage infrastructure inside AWS. So while we help our industry build alternatives to AWS, we also help them to manage hyperscale infrastructure. Our ethos is that workloads should work where they work best. Sometimes that is on-prem, sometimes it’s in co-lo or a hoster’s DC … and sometimes it’s in AWS. We want to be there to help ensure fluidity across all infrastructure and platforms. 

Over the years OnApp has moved from serving the long tail, the SME/SoHo hosters, to mostly dealing with large MSPs and Telcos. That has been great for us, it has made us profitable and given us the solidity to invest in new products and services. However, in the last year or so (especially after selling SolusVM and spinning out Sunlight) I feel that our positioning has moved too far from our original hoster focus – and I am currently looking at launching a new range of products that will lower entry barriers, both price-wise and technically. I want OnApp to be the ideal solution for both the small hoster and the large Telco. Personally, I feel very much at home at LET, it’s just been a challenge to build products that would be financially viable for that very price focused crowd…Watch this space!

SolusVM is heavily used by many providers who participate on LowEndBox. You acquired Solus in 2014, and then sold it to Plesk in 2018. Has Solus’s day passed? Has the industry moved on to better technology?

NO – Solus was (and is) a great software stack. Phil (original SolusVM founder/developer) and his guys managed to *almost* own an entire segment and I honestly feel that without Solus the LET scene would look a bit different. 

We did not buy it to sell it, we wanted Phil to stay on board and get V2 out – but honestly it proved almost impossible. A complete rewrite seemed like the only option and we simply could not find the resources to do that internally.

We were not 100% convinced we should sell, but eventually we got an offer (all in, almost 10x our investment) that was too good a deal to walk away from. Webpros (cPanel/Plesk/etc) is a good home for Solus, and I know Phil has stayed on board and they are doing cool stuff with the brand now. 

Do you think the hosting software industry is going through a period of consolidation? I’m thinking of acquisitions like SolusVM, Plesk, cPanel, etc.

Looking at the software vendors there really are not that many sizeable players left … DirectAdmin, OnApp, Proxmox, CloudLinux and a handful of other players – the rest have been picked up or gone away. Not sure how much consolidation there is left? 

I think new players will enter the market and disrupt what us old timers are doing. At OnApp we are cool with that, it keeps us on our toes and ensures that we always think about product development and keep focusing on creating value for our partners. 

There are areas of our industry that are honestly kind of dated. Like control panels, or billing systems – I think someone should rethink how servers are managed and how customers are billed. It’s all a bit clunky and not very slick. 

At OnApp we are always looking for cool technology to partner with or even acquire – please read this as an invitation to contact me (d@onapp.com) directly if you’d like to work with us.

On your website, it says OnApp is “building a greener cloud”. Green IT is getting a lot of discussion because of climate change. You’re a software company, so how does OnApp make the cloud greener?

From an environmental perspective our industry is among the worst of the worst. In fact our CO2 footprint is worse than the entire aviation industry combined. Since OnApp launched we’ve built thousands of clouds for hosts and other service providers,  so we’re very much part of the problem. 

Last year we decided to take a stand, and do our part. So we’ve committed to offset the CO2 emissions from all of the physical compute resources being managed with OnApp software. Some of that infrastructure is our own – our development, test and demo clouds – but the majority belongs to our customers. So if your cloud runs on OnApp, your company is already a little bit greener. We base the offset calculation on how many active compute nodes there are, across all OnApp clouds, and it takes into account the type of hardware being used, published energy consumption numbers for different hardware types, average utilization rates, and carbon offset conversion ratios in different countries. It was a fairly complicated process, but we wanted to make it so that anyone running OnApp can safely say that their clouds are ‘green’ and CO2 neutral. It’s important. 

What’s your company culture like inside?

It’s an interesting question, especially during Covid19. OnApp is my 6th startup, and I guess you learn a bit from every organisation you build. From that perspective I think that UK2 really showed me the way. There I built segments/domains for each brand and/or product and ensured that the people heading up those domains were fully autonomous, and felt ownership. My role (as CEO) was to facilitate their work, ensure that they had the resources (cash, support systems and infrastructure) available to pursue their strategy. That worked exceptionally well and created a very forward thinking company with a lot of flexibility and agility given the size of the team (almost 300 at the time I sold). 

I’ve tried to replicate the same structure at OnApp, but it is admittedly harder. It’s not possible to fully segment out parts of our product (like CDN, Storage, Networking etc.) and give them a level of autonomy – God knows I’ve tried. At OnApp we have to think coherently and look holistically at every release to ensure that the user experience, APIs and overall quality of the release is on par and aligned across the platform. 

From a company culture perspective that does create challenges as we are spread across three continents. Our main development hub is in Lviv, Ukraine – some of our smartest guys are there. A lot of the key people there have been with me for almost 10 years now and without them OnApp would not exist.

Our CDN developers are based in Kuala Lumpur: we acquired Aflexi some years ago and kept the office and team there. Again, the team there are probably among the top CDN specialists in the world. 

Finally we have our HQ in London, UK. This is where management, some sales and marketing and parts of the company administration is located. In total we are only 20-25 people in the UK. We also have a bunch of people spread around the world in other locations incl. USA, Thailand, Russia etc. 

Each of those locations have a different view on how teams should be run, like, what it means to be a manager … or at a very practical level, how products are built. In Ukraine they are very strict and almost militant when it comes to PRDs, SCRUM etc – where it’s quite a bit different in Kuala Lumpur, here ‘user stories’ are the predominant planning tool. Combining the two can be a challenge and we’ve had many robust discussions on that topic over the years.

Across all offices I try to keep a very free, open line of communication. Our management layers are few and flat and we have anonymous ways of constantly capturing team feedback. We use a system called Peakon for that – and it works great.

I just realised I didn’t mention how Covid19 have changed our company culture – well, it has …  in a big way. We’ve gotten much better at communicating, and keeping each other up to date. And ‘work-from-home’ (that I frowned upon before) is now the new normal – and it just works. 

Your products are both for public and on-prem clouds. How do you see clouds evolving? When the buzzword came along, there was a vision that within a few years, private data centers would be a thing of the past, but they’re still alive and growing.

It’s all coming together into one pool of infrastructure. On-prem, off-prem, public, private, IaaS, PaaS, SaaS etc etc – those terms are much less interesting today than just a few years ago. Now it’s all about fluidity of workloads, ensuring that devops teams across our customers or partners can move workloads from location a to b … from on-prem to off-prem … from VMware to KVM … from AWS to local infrastructure etc etc.

It’s where our development efforts have been for the last number of years … and also why we built Cloud.net.

Tell us about your latest offering, cloud.net.

So, Cloud.net is the next step towards our goal of simplifying infrastructure for service providers, and enabling that fluidity I was just talking about. It’s a way for you to spin up a complete public or private cloud without CAPEX on hardware, and without long-term OPEX commitment either. We provide the whole cloud environment as a managed service, the OnApp cloud platform and the underlying hardware, and it comes with 24×7 support and 5×9 SLAs.  And the pricing is really simple, just pay month to month.

It’s a cloud in a box, easy to deploy and instantly available.

The ultimate goal with Cloud.net is to provide Cloud-as-a-Service, and we have much more to come on that front. I basically feel that it should be much simpler and cheaper to set up clouds. We are not quite there with Cloud.net yet, but keep an eye out for what’s to come :)  

Every time you turn around today, someone is slapping the term “AI” on a product. For cloud hosting, how do you see AI playing a role – both today and in the near-term future?

Oh man – that’s one of my pet peeves. AI will obviously be huge, it will be awesome and it will change the way we all do stuff on a day to day basis. However we are not there yet, and most of what people call AI is simply If-Then statements that do not get smarter with time or learn from past processes. 

Now, there is an AI angle in our industry, for sure. I feel that workload locations would be a great place to start. Workloads don’t care where they are, they just want to work. And AI could help moving workloads around to ensure they are optimally placed – both from a financial, security and performance perspective. We have discussed this a lot internally at OnApp and we will bring out tools to make that possible.

What do you enjoy most about your role? What do you find most difficult?

I am a product guy at heart. I love getting my hands dirty and building out new product strategies and rethinking the way we all do business. It’s harder at a maturing company like OnApp. Things take time and sometimes I get frustrated in the time to market, long priority lists and all the different priorities that we just have to respect before focusing on a specific development or new initiative.

I do a lot of angel investments (through www.kickass.capital) and I envy the startups that I work with there. It’s just much easier at their stage. On the flip side, OnApp is financially solid and that obviously brings a great deal of comfort as well.

Both personally and professionally, what guiding principles ground you?

My kids (14+16 years old) and I often say ‘all you can do, is all you can do’. I feel, both personally and professionally, that we should strive to do our utmost, but also know our limitations so we can build a network – both personally and professionally – that can help us succeed where we have shortcomings. All you can do… is all you can do. I love it – it’s what I expect of myself, and people around me. 

I'm Andrew, techno polymath and long-time LowEndTalk community Moderator. My technical interests include all things Unix, perl, python, shell scripting, and relational database systems. I enjoy writing technical articles here on LowEndBox to help people get more out of their VPSes.

6 Comments

  1. someone:

    Great read and very interesting story. Thank you for doing this!

    As for SolusVM, it “was” and no longer “is” great back when it was still relevant. Don’t get me wrong, it works, but no one put in the effort to update it. I’m not talking about V2, just simple feature updates to keep it relevant is more than enough for many low end providers.

    When it comes to the lower end sector, I feel Virtualizor is the best option right now in terms of features and its great value.

    September 2, 2020 @ 2:50 am | Reply
    • I know – it was super frustrating … but really impossible for us to get V2 done without a complete re-write. We even considered doing an ‘OnApp-light’ and just call that SolusVM V2.

      September 16, 2020 @ 5:48 am | Reply
  2. Thanks for taking part in our interview series Ditlev!

    September 2, 2020 @ 9:52 am | Reply
  3. Another great interview! Good work, excited to read more of these.

    September 2, 2020 @ 1:49 pm | Reply
  4. Having used SolusVM from long and a fanboi as well, I was disappointed when there was no SolusVM v2. The false promises of releasing the v2 every year further broke me :( This made me to shift to Virtualizor and now I’m a bit reluctant to try Solus.io
    Keeping aside the internal reasons, OnApp could’ve again reigned as the market leader if they would have released the SolusVM v2.
    You drowned a lot of ships of hope, my friend.
    Nice article and great insights.

    September 12, 2020 @ 7:58 am | Reply

Leave a Reply

Some notes on commenting on LowEndBox:

  • Do not use LowEndBox for support issues. Go to your hosting provider and issue a ticket there. Coming here saying "my VPS is down, what do I do?!" will only have your comments removed.
  • Akismet is used for spam detection. Some comments may be held temporarily for manual approval.
  • Use <pre>...</pre> to quote the output from your terminal/console, or consider using a pastebin service.

Your email address will not be published. Required fields are marked *