New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
The OnApp $500 tag is a minimum spend not the price for 48 cores. At scale the cost is a tiny fraction equivalent to a servers power cost. (Sub $50-60)
It does when we looked at it but OnApp doesn't look great inside once the UI is taken away for example this was a good read when looking at such "cloud" platforms:
http://www.webhostingtalk.com/showthread.php?t=1531673&page=4&p=9613351#post9613351
I do believe as suggested above Proxmox is ever so looking more popular especially on the KVM side. 3.x (w/ OpenVZ) will stop being supported in April so not sure what will happen then, either providers will live on edge or move to another platform.
I have never really liked any sort of "revnue sharing" models. Per core model isn't terribly bad but it purely depends on who you are targetting. This won't attract a lot of hosts here for example selling <$20 VMs with little margin with you taking a further 10% off.
This is very insightful, thanks for the actual information. This is the main purpose of asking for feedback!
How do you think a minified version of our product would work in this market? Being absolutely affordable too! Probably will still support the fundamentals -
Hourly Billing
Torpedos (VM's)
On-Demand Infrastructure (Clients deploy & destroy there services)
But won't support the more sophisticated features such as PaaS, AutoScaling, H/A! Unless the user wants to upgrade to the enterprise version which can stand just fine at 10% -> 2% as mentioned previously?
I think it's a waste of your likely limited developer hours.
Virtualizor is a good example to look at. They undercut SolusVM on the initial license and are likely willing to cut deep discounts if you're buying volume. Even still, there is no large scale host using them and the few that do usually run back out the door because the owners have ADHD/tourettes and can't make up their bloody minds.
Hostguard also tried to come to the market a couple times now and has failed to get a real grab at things (though to be fair they had some deployment issues to fix first). They had some $5/month/node licensing and even then people wouldn't budge.
You have to compete with more or less ~free from the Proxmox side of things. While the modules are way uglier than your platform, some bootstrap here/there and a weekend of time and they'll be set. While SolusVM is a little more complete since it has RDNS & DNS hosting built right in, but still, I think people are willing to lose that, or can't use those features anyway due to their upstreams not doing delegations/them not caring enough.
Honestly, anyone that would be interested in your offering is already trying to home roll a platform. The majority of hosts you see on these forums are maybe a half dozen nodes deep, and many of them offer plans that are loss-leaders in hopes of growing brand recognition, if you're now telling them they will have to eat even more just for the panel, it gets rough.
99% of the hosts on here can't offer HA in any shape/form since they are leasing dedicated servers and they aren't going to be able to budget in a $500/m big ass storage node to act as a SAN.
You're simply barking up the wrong market/tree and unless you have VC behind you to cover your admin & development time, I would seriously sit down and light any business plans that targets the budget market on fire, as they'll be more useful as kindling than anything profitable.
Francisco
Our platform takes care of this, you don't need SAN's for H/A with our platform. This is one of the selling points of our software - How much cheaper do things become when you don't need SAN?
DET.io v1.0 will have Route 53 integrated for automatic hostname assignment, and rDNS. DNS will be added in a later stage!
We are building automated deployment scripts, OpenNebula is already easy to install! We have a script to install both the host and front-end of open nebula right now. Just looking to hook it up further with the platform so it's all automated.
The solution is built, I don't see how we can't strip it out for budget companies in order to save value in the product while still offering it around - it doesn't hurt is to increase deployments (If this version ever comes to life, it won't be unbranded however)! We aren't looking to eat up clients funds but to help them grow so we can have a bigger piece of the pie.
Budget companies may bring a lot of exposure to the platform, which is something we need - Backlinks will help from these deployments too.
How much does each core cost after the 48 limit? I haven't talked with their sales for a while (Last time I was in touch was years ago, and I just noticed there scaling might have gotten a bit cheaper?)
Not all hosts are willing to do per-customer VLAN's, meaning when an HA happens the IP's won't be routed over. Far as I can tell most of the large shops have each dedicated server on its own VLAN and only do a customer VLAN if the customer either buys enough or pays a cost to do so, due to it usually just being easier to move the customer into a single spot and have them on their own switch if they have enough servers.
Great, how do you deal with hosts that won't delegate the records? With spam being such a huge issue these days there's many hosts that insist on doing manual RDNS activation.
Unless you're going to offer the platform free for a year+ like OnApp, I feel you guys are going to have a really poor turn out and it's going to really kill your moral.
Like I said, the hosting communities had a real chance to get a complete VPS/Cloud platform, all they had to do was each chip in a bit of cash. They didn't and now they get nothing.
You should think of much of the hosting communities as a real-life telling of "The little red hen".
Each of the big brands out there could've chipped in $1k each and it would've been funded in an afternoon and who cares if everyone else gets access to it, you're at least not on the hook for gobs of cash to SolusVM for a lackluster, exploitable, platform.
EDIT - reorganizing things a bit.
Francisco
This is taken care off by Route53, after VM Creation the owner is noted to use the hostname instead of the IP. This is more mandatory for Autoscaling however - Since VPS's are migrated over from one node to another when the node starts to bleed resources. There is prediction algorithims in place that allow the largest VM to migrate without loss of power (It takes 10 minutes for a migration on a public network, way less on private ) the VM is synchronized over to another node and then once the VM is deployed on the new server the old one is switched off after route53 changes IP's. The IP changes are only necessary for cheaper deployments - Most of our bigger clients don't need this.
In a latter version where DNS is also hosted, users that use our nameservers will have there domain forwarded always to the specific server - Even if IP's change.
Then again this is a ready product that can offer value for a business, a free version would probably not exist - Simply because of the sheer amount of support you have to deal with. I'd probably opt in for a similar price to SolusVM with a demo explaining how they can migrate from SolusVM within the hour and how it adds value.
The hosting community can be seriously dramatic at points too. But then again everyone here wants to make the others look bad haha!
If it's already complete then I guess you got the heavy lifting done.
You could undercut Solus, but I still think it'll be a hard sell.
Nothing more I can add this than good luck
Francisco
Thanks for the feedback, helped me construct a decision on how I can handle over sensitive LET users while keeping everyone happy!
"Get the fuck out of dodge"
Francisco
Nah never would, in my experience it got used way more than I thought it would and being able to set own rDNS is nice. Getting delegation from dedi provider is simply a case of using what I like to call "VPS host friendly" providers who know what VPS hosts need and what a good one looks like (QuickPacket for example are very good at this).
SolusVM's DNS is one of the features that actually works great, can be scaled easily and is very easy to setup, if you ignore the lack of sometimes confusing documentation that is.
It's a great feature and saves ticket volume, but it's still not super common short of the dedicated hosts that heavily market to VPS companies.
Francisco
Beware SolusVM 2, we all know it will support all these crazy features - right
This is OpenSources biggest opportunity and biggest failing. At some point members of a project have differing views and instead of resolving them a project forks and after x number of forks none of them have the inputs to deliver and they all fail, quickly or slowly...
They won't let SolusVM undermine it's sister product (OnApp) - According to Ditlev (CEO of OnApp) - They aren't introducing any cloud FEATURES in SolusVM. That already puts us on an advantage.
I know I was just joking.. because V2 has been in production forever :P
Make it, if it stands up then it will work, the absolute bottom line here is if you have faith.... make it, but don't expect much excitement from anyone until we can see the finished product.
Personally I would pay 10%, I think its an idiotic way to price things from a business perspective though.
If you can deliver a truly next generation panel, that really does offer more, that can import everything from solusvm and whmcs at the click of a button... then just do it.
But keep in mind that 90% of end users get the initial root password and IP and never login to the panel even once, so your focus is on 10% of the entire user base to begin with, in order to entice people that already know what they are doing you really better have some bells and whistles.
Alright, so here's my thought on this - 10% of total income is ridiculous.
Let's say I hypothetically would want to use your panel to start a company offering web hosting and VPS hosting (this seems to be the most common path taken by LowEndProviders)... or add VPS hosting to an already existing company offering web hosting. WHMCS would be the obvious choice for a billing system. Obviously your virtualization system would have some sort of WHMCS module, so in theory, your module would have access to WHMCS to report total income from only VPS services/invoices, and report that back to your own servers. IMO that's a security risk to the hosting companies and an invasion of privacy.
Personally, I prefer a per-node approach like SolusVM. Now, why would somebody choose your solution over SolusVM when SolusVM is owned by a huge company that is guaranteed to be here tomorrow, and it is proven to be at least somewhat functional, stable, and (for the most part) secure for many years? An established provider would most likely not make the switch. A new provider would go with a proven solution, unless your product is super cheap or even free. What I'm trying to say is the only way your product is going to grow quickly is to significantly undercut SolusVM's pricing.
Here's my input if you want your product to gain traction in the LowEndMarket... make the product completely free for up to 3 nodes, and make it $5/month per node after that. That way, new providers with no up front investment will all choose your product, and then moving forward once they grow and need to pay for it, it'll be half the cost of SolusVM, and cheaper than Virtualizor.
We have a billing system that reports the billing directly from our Software - WHMCS integration is what is used to Add Credits & Allow for Admin management of accounts (Accounts that get suspended on WHMCS get suspended on our Software) etc... Until the AdminUI is completed (Which is after V1.0 Is released). Part of the support modules also have WHMCS as an option - So you can integrate that directly instead of using 3rd party.
I understand the concern for this, and as mentioned earlier we have decided to build a stripped down version that will allow us to significantly undercut pricing while maintaining our software's value! Worry not, the software will be by minimum-standards competing with DigitalOcean's panel (Which is by far superior to SolusVM).
Thanks for the feedback, will take it into account!
You have seen the admin side have you?
Honestly the client side needs boot/reboot/shutdown/reinstall and some stats... that's it, a VPS is not a control panel, 90% of users never log in, if your focus is on the end user you have completely missed your market.
Virtualizor has a far more features as an end user control panel than solusvm, they are gaining zero traction because the admin side is absolute bollocks and completely fucked at the best of times.
OnAPP's admin side is leaps beyond solusvm, end user side... meh it does the job and on scale not starting price it would be cheaper than your solution.
If you can make an admin panel that saves me time over solusvm and adds value to me as the person running the show I am in with both feet, show me the beta... seriously, show it to me.
This is not a concept, but a product that's near completion
Our website will be updated over the next upcoming weeks but right now you can see screenshots on our website 
As previously said, 10% is only a start for the enterprise version. Since we want to ATTRACT big hosts to use our enterprise version while maintaining the value of our software, most big hosts don't mind paying the 2% that the pricing scales down to!
The added features to the panel come handy, things like Hourly Billing - High Availability - Snapshot System - ETC. Come in really handy even when a user doesn't use the panel. Not only will the host be able to import servers from SolusVM but the user will have access to this @ 2.0 so our clients will have competitive advantage over other peers because if they are able to offer some sweet reliable servers, some users might enjoy the added features over Solus and migrate their services.
The virtualization platform handles the administration of VM's, WHMCS is only used for the administration of accounts. Both of these will be combined in the AdminUI with a more intuitive interface but EVERYTHING can be done right now utilizing OpenNebula's Admin Panel & WHMCS.
Well it sounds good, proof is in the pudding as they say.
We can get you a demo running early Wednesday, how does that sound? It won't have WHMCS integrated yet but that's due with v1.0 release which is scheduled for 26th/02/2016. We are doing a lot of testing right now, 90% of the platform is built.
The major things that don't EXIST yet as actual functionality (They are in design process) is only the following:
Creating Snapshots - Rebuilding VM's from Snapshots - Automating Snapshots with Scheduler - Downloading Snapshots (Will also be available if a user is suspended) (either v1.0 or v1.1/2 will support this)
Building from ISO. (either v1.0 or v1.1/2 will support this)
Auto Scaling and Implementing our Anti-Bleed Mechanism. (Scheduled with high priority after v1.2)
Load Balancing (Scheduled with high priority after v1.4).
Email me at [email protected] so we can schedule a demo.
For now you can have a look at these screenshots:
Are you also selling/renting torpedos to people or are you just looking to hook people/providers on the platform?
That looks really quite good, but I'd have to say paying a 10% commission doesn't sound too appealing; regardless of the visual appeal. I'd much rather pay a SolusVM/Virtualizor style monthly per node fee, if it can scale with the amount of licenses, all the better. But that's not necessary. I'd be happy paying $10 per license for 10 licenses, just as I'd be happy paying $10 for 1 license.
Also, is the name "Torpedo" customizable in a config file/admin area or similar?
Torpedos is what we name our VM's, because our templates allow our systems to communicate directly with the VM. For further functionality - Like better log systems and monitoring data. Enterprise version will allow the full modification of templates. We haven't set a specific pricing model but enterprises will have the option to chose between three different pricing models (Setup + Support fees - Scales the best/Per Host specs ($X/xGB ram)/% Revenue Share (Which was for the sole purpose of hosts that are in the LowEnd market - Since their pricing model is lower, licensing won't hurt).
We built the platform for hosts to use, visit us at http://det.io to learn more. That's why the screenshots come with a custom logo 'CloudHost' (Not a company, if it is a company then this is just a dummy logo that does not represent CloudHost) as you can personalize it.
When there is a version I can install I will test it.