Cluster Águia Pescadora [Nodes: 3][CPU: 12][RAM: 24 GB][Disk: 600 GB][Cost: (14.97 EUR)/month]

[fititnt]

Wikipedia: Pandion haliaetus | Image source: pxhere

• Servidores Águia Pescadora fititnt/cplp-aiops#5 ^(Portuguese)

---

TODO: add basic details about this Cluster on next weeks

[fititnt]

The cluster was reduced to Cluster Águia Pescadora [Nodes: 3][CPU: 14][RAM: 32 GB][Disk: 800 GB][Cost: (18.97 EUR)/month] (old size: Cluster Águia Pescadora [Nodes: 5][CPU: 17][RAM: 37.5 GB][Disk: 870 GB][Cost: (8.49 USD + 18.97 EUR)/month])

At this moment, we have active

• VPS Águia Pescadora Delta [6vCPU / 16GB RAM / 400GB] [8.99 EUR/month (*)] VPS Águia Pescadora Delta [6vCPU / 16GB RAM / 400GB] [8.99 EUR/month (*)] #80
• VPS Águia Pescadora Echo [4vCPU / 8GB RAM / 200GB] [4.99 EUR/month (*)] VPS Águia Pescadora Echo [4vCPU / 8GB RAM / 200GB] [4.99 EUR/month (*)] #86
• VPS Águia Pescadora Foxtrot [4vCPU / 8GB RAM / 200GB] [4.99 EUR/month (*)] VPS Águia Pescadora Foxtrot [4vCPU / 8GB RAM / 200GB] [4.99 EUR/month (*)] #87

And two active dedicated repositories

• https://github.com/EticaAI/aguia-pescadora
  • https://github.com/EticaAI/aguia-pescadora-ansible-playbooks
  • https://github.com/EticaAI/ajuda.etica.dev

The original project AIOps para Comunidade dos Países de Língua Portuguesa, 2019/01 (https://github.com/fititnt/cplp-aiops) will have some of it's ideas used for later. Maybe not in the original idea of the Alpha.

This new approach can take some time to a working cluster, not because it is not possible to have a big one up and running. Get very good prices (like the ones from the provider of Delta/Echo/Foxtrot, even if not sponsored for anyone else) was impressive to get things going, and I know that eventually we will have some working. But, at least for the year of 2019, and the state of the use of containers and awareness of how to use there is lack of documentation and the average developer already does not know very well how to use containers.

Maybe even if things work well, instead of a big cluster, we will eventually have smaller ones and focus on first have documented strategies on how to move apps, disaster recovery, etc than focus on high availability on more short term. I recently found that there is even a term called "Kubesprawl" about have "many clusters" instead of "one big shared", so what I'm thinking here is not even that strange.

[fititnt]

We will retire the Delta VPS and immediately we're adding Golf.

Golf has the same price and sizes than Echo and Foxtrot. Since most of setup is Ansible automated, for redundancy propuses is better to have 2x [4vCPU / 8GB RAM / 200GB] than just 1 x [6vCPU / 16GB RAM / 400GB]. The smaller ones already comes with one IPv4 (and 4+4 vCPUs, vs 6 vCPUs).

Another important fact is that even the sizes of Echo/Foxtrot/Golf alone already are good (we even could reduce more, but these VPSs already are the cheapest package on the cloud provider from Germany we got).

[fititnt]

Updated cluster size to Cluster Águia Pescadora [Nodes: 3][CPU: 12][RAM: 24 GB][Disk: 600 GB][Cost: (14.97 EUR)/month].

The changed was mentioned at the last comment (#82 (comment)) but we forgot to document that the Delta was closed

[fititnt]

See #92 (comment)

Addendum

At least two machines are still on, and one of them not used for projects at all. Everything still with history, but we migth rebuild it later with either another name or new versions.