Skip to content

cirrus: halt the local fallback deployment - #144

Open
cboettig wants to merge 1 commit into
mainfrom
halt-local-fallback
Open

cboettig wants to merge 1 commit into
mainfrom
halt-local-fallback

Conversation

@cboettig

Copy link
Copy Markdown
Member

Scales the cirrus deployment to 0, matching what is now live in the cluster.

Why

This copy existed as a fallback for if NRP went down. On reflection that is not a good reason to keep a second deployment running: it holds a public hostname and needs upkeep, in exchange for standby capacity nobody was using.

It was also genuinely idle. Over 24 h it logged 5 requests, all crawler traffic:

GET /            200
GET /favicon.ico 404
GET /robots.txt  404

No API calls at all.

Scope

Only replicas changes. The Service, Ingress and the rest of the manifest are untouched and still correct, so restoring is setting replicas: 2 and re-applying — the comment in the file says so.

The NRP-facing deployment (the actual purpose of this proxy: GLEN models plus CORS and request logging for web access) is unaffected.

Scaled to 0. It was kept as a fallback for NRP outages, but served no real
traffic -- 5 requests in 24h, all crawler hits on /, favicon.ico and
robots.txt, no API calls -- while holding a public hostname and needing
upkeep. Set replicas back to 2 to restore.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant