Official Documentation for Self-Hosted Deployments
I read these docs. Is it possible to run everything on just 1 system? I want to evaluate Yadle, but do not want to find 2 systems to install on. Also, can I run all of this on a VM or even on AWS?
Hi @debers - Welcome to the Yadle community!
Great question as we know wrangling resources for new software deployments can be tricky.
While it’s acceptable to run all Yadle components on 1 system, don’t expect to achieve the same performance as a properly spec’d split system deployment (1 server node + 1 agent node).
At a minimum, to run Yadle on a single system, we recommend 4 cores and 32 GB ram.
The system you choose to install Yadle on is up to you. As long as it is running a modern linux distribution (we recommend Ubuntu 20.04), you can run Yadle on a local workstation or VM, or on a VM from any cloud provider including AWS & GCP.
If you’d like to get started with a Yadle eval, fill out this form and we’ll get you going.
Thank you Erik. Let’s go the other way - is there an upper limit to how many systems I can run? If I use in production, how many systems to continuously handled files on my 2P storage?
Can the systems running Yadle be in multiple locations - like LA, London, Montreal - and all work together? Or is each office need to have it’s own Yadle?
Any number of Yadle agents can be deployed across separate geographical locations, each connecting to a single Yadle server for all users in your organization to access.
The Yadle microservices, which perform post scanning operations such as thumbnailing, tagging, metadata extraction/normalization, can be run on any number of systems as well. This flexibility is made possible since all Yadle services are deployed as Docker containers.
Note that the agent-side microservices require read access to files and usually run alongside a corresponding agent.
Here’s one possible deployment to illustrate the flexibility:
System 1:
- Yadle Server
- 12 Core 64GB
- Running in LA
Systems 2,3,4:
- Yadle Agents
- 8 Core 32GB
- Deployed in LA, Montreal, and London
System 5:
- Yadle Microservices (server-side)
- 4 Core 16GB RAM
- CUDA Enabled GPU for local AI analysis
- Running all server-side microservices
- Deployed in LA
System 6,7,8:
- Yadle Microservices (agent-side)
- 4 Core 64GB RAM
- Running 3D thumbnailing and metadata microservices
- Deployed in LA, Montreal, and London
System 9,10,11:
- Yadle Microservices (agent-side)
- 8 Core 32GB RAM
- Running all image and video thumbnailing microservices
- Deployed in LA, Montreal, and London
Thank you for the details. Can I run Yadle (everything) on my laptop? My laptop is new I9 and has 64G of memory.
I am thinking I would run on my laptop to explore how it works before getting IT department involved is setting on a more.
A laptop with a new i9 CPU (12+ cores) and 64GB RAM is more than adequate to run a complete Yadle PoC (Proof of Concept).
You can install and run all of the Yadle components on it. Performance will vary depending on location of files to be scanned (eg. local, network, external volume) as well as if the laptop will also be used for other day to day activities besides Yadle.
We’d love to get you started with a PoC. Just fill out this form mentioning you’d like to start a PoC and we’ll get you going.