Frequently Asked Questions
Q: Why did you rename Jidoteki?
A: Despite its uniqueness, 'Jidoteki' has always felt like an awkward name when used with non-Japanese speakers.
Q: Why did you choose the name On-Prem?
A: We wanted a name that exactly reflects what we do, while avoiding the issues of the 'Jidoteki' name.
Q: Why did you cancel the Jidoteki SaaS?
A: We transitioned to a distributable OVA and Managed Service due to customer demand. It also reduced our hosting costs, and free'd us from being on-call 24/7.
Q: Is the On-Prem Meta appliance really necessary?
A: The 'Setup' only includes the initial OVA build and minor tools. Meta allows you to DIY with your own continous integration tools. We also offer Support if you want to offload updates and weird edge-cases to our team of experts.
Q: What is the difference between a rootfs and an OVA?
A: The rootfs contains all of your app’s software dependencies. The OVA contains the boot disk, which contains the rootfs, unmodified OS, and vanilla Linux kernel.
Q: What’s the typical size of a final OVA and Update Package?
A: It depends on the size of your application and dependencies. Our base OVA is ~35MB, and Update packages are typically a few hundred KB to a few hundred MB. It's impossible to predict before the first build.
Q: How do you handle updates when there’s no internet?
A: The OS runs entirely in memory, which allows the system's OS, kernel, rootfs, and your app to be updated atomically without affecting the running system. Update Packages can be generated to perform these updates offline.
Q: How do you handle database migrations?
A: DB migrations run on boot, so you can guarantee they will only run against a specific version of your application. It always seems like the most difficult issue, but it's actually our easiest.
Q: Can you revert a failed migration?
A: Of course! On-Prem virtual appliances ship with LVM by default, and allow an LVM snapshot to be taken before performing a migration. If it fails, just revert to a working state.
Q: Can you do hardware builds?
A: Yes! We've deployed installations to Beagle Bone Black, Raspberry Pi, and ROCK64. The final system is almost identical to a virtual appliance, except designed for arm32 and arm64.
Q: How is Meta different from VeeWee/Packer/Vagrant
A: Meta is a stable solution which doesn't introduce backwards incompatibilities or breaking changes. It just works. There's no learning curve or special tools to compile/install. You won't spend weeks editing complex nested YAML files.
Q: How is Meta different from Docker/Moby/LinuxKit
A: Meta is designed for building virtual appliances which run directly on a hypervisor. It doesn't force the use of containers and unstable technologies. It's extremely simple to use and you can start building virtual appliances within seconds of downloading it.
Q: Do you support Docker?
A: Yes! We have customers deploying production virtual appliances with Docker and pre-cooked container images built to their specifications.
Q: Do you support AWS (Amazon Web Services)?
A: Yes! Our Meta OVA can run on EC2, therefore our customers' OVAs can also run on EC2. Our Meta OVA builds OVAs which can be deployed on a typical hypervisor, or as an Amazon AMI.
Q: How is this better than other paid solutions?
A: We build everything on the premise of being 100% Open Source. You'll never be locked-in to our tools or platform, and you have the freedom to modify your appliance however you want, with or without our intervention.
Q: Everyone is going to the cloud/containers/serverless!
A: There are countless companies unable and/or unwilling to use cloud solutions due to the nature of their business. Those companies can greatly benefit from modern and stable web-based software running locally.
Q: Is On-Prem Meta going to be open sourced?
A: It's already 95% open source, but yes, we plan to fully open source it once we've completed writing proper documentation and tests for all our code. Give us another year or two ;)
Q: What other benefit(s) do you provide?
A: You can focus on your software instead of this 'appliance' thing. You'll incur significant cost savings by refocusing your engineers on your core business, and you'll generate more revenue thanks to how quickly you can release updates and new features.
Q: How long does it take to get started?
A: It depends on your stack and application. Typical LAMP/LEMP/MEAN stacks can be ready within a few hours, but on average you can expect 2 to 4 weeks of iteration to get a final production-ready appliance.
Q: What if we have a big dreadful bash/perl/php script with an embedded rpm?
A: No worries, we'll convert it to Lisp ;)
Q: How does the Meta License work?
A: Our Meta appliance comes with a perpetual license. You pay a yearly subscription for updates, bug fixes, and support with your installation(s), but you're free to continue using it even if you don't renew your license.
Q: What does the Meta License allow?
A: You are provided unlimited users and can launch as many instances as you want (eg: if you have multiple offices or multiple teams). You're free to login and view the source code. You can even edit the code (at your own risk) and appliance if you know what you're doing. Of course, changes may be wiped by our updates. The Meta License does not allow redistribution or resale of our source code or Meta OVA.