Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Estimate cloud costs for 2023 #788

Closed
pnorman opened this issue Nov 21, 2022 · 3 comments
Closed

Estimate cloud costs for 2023 #788

pnorman opened this issue Nov 21, 2022 · 3 comments
Labels
location:aws Services hosted on AWS planning

Comments

@pnorman
Copy link
Collaborator

pnorman commented Nov 21, 2022

The cloud costs are the only recurring costs that vary significantly, so they should be worked out separately from other recurring costs as they require future planning.

@pnorman pnorman added the location:aws Services hosted on AWS label Nov 21, 2022
@pnorman pnorman added this to the 2023 OSMF budget milestone Nov 21, 2022
@Firefishy
Copy link
Member

Need to add Azure costs too, Azure is used at the moment for Translation services for community.osm.org
Current costs are expected to be £26 per month. Cost will increase as adoption of community.osm.org site grows.

@pnorman
Copy link
Collaborator Author

pnorman commented Dec 16, 2022

I've budgeted Azure under subscriptions since it's so small compared to other hosting costs, and it's used for one specific service.

@pnorman
Copy link
Collaborator Author

pnorman commented Dec 16, 2022

So, I had a browse through AWS cost explorer. There's a few things that make budgeting hard.

  1. Since August 12th, Set openstreetmap-fastly-logs S3 retention to 30 days #727 cut our AWS costs in half. This is, of course, a good thing, but it means we're restricted to 3.5 months of data we can use to forecast.
  2. Costs consist of a predictable baseline with spikes for one-off batch operations. These batch operations might tasks like back-populating log tables, copying new files to S3, or ad-hoc of tile logs.
    image
  3. Tile log retention and aggregation #698 is not yet done - this should reduce our long-term costs as we will retain less data, but will shift some costs around in unknown ways.
  4. For some reason, we've been getting credits since Oct 1st equal to our spend. We haven't had any bills to pay since the September bill which came in on October 2nd. This is excluding the render account. Because I don't know the source of this, I'm excluding these credits from the cost analysis.
  5. Part of our S3 usage is backups, which we do not access on a regular basis. If we ever needed them, we would have significant bandwidth costs to download them.

Given all of these constraints, I fit the weekly data to a curve and got a cost of 26100 USD. This matches what I get for monthly data. If I used a linear fit to the last three months, I got 21k USD.

@pnorman pnorman closed this as completed Dec 16, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
location:aws Services hosted on AWS planning
Projects
None yet
Development

No branches or pull requests

3 participants