conjurer: add production ('deploy') bot + data seeding/backup

A second Conjurer bot alongside the test one, sharing librarian/musician/
radio and the API key, differing only in:
 * Discord token from the deploy-conjurer-netrc secret,
 * distinct names/labels (deploy-bot) and NodePort 32443,
 * /data on an NFS export (RWX) instead of a block PVC - so config/state
   can be uploaded while it runs (copy onto the share) and backed up
   concurrently.

deploy-bot-backup: a daily CronJob that mirrors /data and keeps 30 days of
dated snapshots of the critical small state (both memories, settings,
accident log, transcripts) on NFS. Production only - the test bot's
amnesia is fine. DEPLOY-BOT.md documents seeding, backup/restore, and the
librarian/musician callback routing (they push to one bot; repoint
CONJURER_MAIN_BOT to :32443 to feed this one).

kubectl kustomize builds cleanly (10 objects).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

This commit is contained in:

migatu

2026-08-03 18:07:53 +02:00

committed by

gitea

parent 71133be104

commit 358356a1b1

4 changed files with 214 additions and 0 deletions

									
										conjurer/kustomization.yaml
									
		+2
		
												View File
												
				@@ -4,6 +4,8 @@ resources:

				  - namespace.yaml

				  - librarian.yaml

				  - bot.yaml

				  - deploy-bot.yaml

				  - deploy-bot-backup.yaml

				images:

				  - name: gitea.czernobog.pl/gitea/conjurer-librarian

				    newTag: ac16b77f