Monitoring and automated deployment of subaccount nodes

[leandrosardi]

OPERATIONS

• create/remove nodes automatically.
• install environment in new nodes --> AMI / How to deal with Ubuntu version retirement?
• setup subdomain pointing to that new node.
• setup LetsEncrypt SSL for one or more domains in that new node.
• deploy new versions on nodes
• monitor parameters thresholds, logs keyword, processes are running, proxies working, watch logfiles, show list of alerts

---

There is a new project: Monitoring servers for a new SaaS.

For servers deployed using TerraForm, we need to monitor:

• Server parameters, with thresholds (you tell me what parameters and treshold should I watch).
• Raising alerts via email when the world "error" appears in a log file.
• Raise an alert if one of the processes lanched by our SaaS is not running.
• Watch log files online.
• Monitor that proxies are working.

If you are interested I can share the full requirement.

---

There is a new project: Auto-Scaling servers for a new SaaS.

Using TerraForm for creating/deleting servers with a specific image.

Also, also we need to setup a new subdomain with SSL for each server.

If you are interested, I can share the full requirement.

---

Please checkout this model:

https://github.com/MassProspecting/docs/raw/main/assets/internal/1-1.png

Technologies:

• Docker for creating images
• Terraform for deploying new servers.
• any other tool for upgrading/downgrading version of our saas in the nodes.

Do you know how to do that?

---

Technologies: Docker, Terraform.

Related documents:

• MassProspecting Architecture
• [MassProspecting Installation](https://github.com/MassProspecting/docs/blob/main/internal/2-installation.md

1. we need to choose a cloud-hosting provider: contabo? aws? hetzner?

2. create images of different type of nodes that we need to work with.

There are 3 differnt images:

• master,
• slave, and
• worker.

The manual procedure for these 3 kind of images are explained here.

Images have databases installed, so they must persist after restarting. If we are docker instances, such instances must persist.

3. request creation/deletion nodes using TerraForm.

@leandrosardi should should be albe to request creation/deletion of nodes from Ruby code, passing the following parameters:

• role: master, slave or worker
• hostname: to assign to the node
• password for the root user
• password for the blackstack user of ubuntu and blackstack user of postgresql

@leandrosardi should be able register created/deleted nodes into our database, with the following informaiton:

• public IP to connect the database and node via SSH

[leandrosardi]

Draft 1

goal:

Perform all operations with no need to access server via SSH

deployment

locally push secrets remotely pull secrets install environment remotely install components remotely deploy db deploy code deploy extensions stop nodes start nodes

monitoring

watch processes running watch node CPU, RAM, disk space watch logs email alert when a keyword happens in logs email alert when CPI RAM or Disk Available Space goes over a threshold. email alert when a proxy stops working