docs(ops): make the cutover runbook agree with the compose file about DEMO_MODE (#196)
Linting / lint (pull_request) Successful in 1m54s
SonarQube Analysis / sonarqube (pull_request) Failing after 19m47s

#190 turned `DEMO_MODE` from a hardcoded compose value into a Portainer stack variable with no default, and updated the compose header. The runbook that actually creates the stack was not updated with it, so the document and the file it deploys have been disagreeing since — in the three places most likely to be read under pressure.

Step 2's list of stack variables to record did not include `DEMO_MODE`, and step 6 says to add the variables from steps 2 and 3. An operator following this literally creates a stack that cannot boot. It is now in the list, with a paragraph of its own: it is the only one of the nine that is not a secret, which is exactly why it is the easy one to skip past.

The troubleshooting section said `DEMO_MODE` was hardcoded and therefore could not be missing, so its absence from the error list proved nothing. That was the most dangerous sentence in the file — an unset `DEMO_MODE` is now the *first* thing to check rather than something to rule out. The line now says it moved and why.

The crash-loop example was the PayPal triple, which cannot occur while `DEMO_MODE` is `true`. The message an operator will actually see during the demo interim — `DEMO_MODE is required and must be exactly 'true' or 'false'` — appeared nowhere in the runbook. Both forms are shown now, in the order they are likely to be hit.

Added what neither document said: Compose only *warns* about an unset variable and deploys anyway. In Portainer's stack UI that warning is easy to miss, and the container then crash-loops under `restart: unless-stopped` — loud in the log, invisible in a glance at the stack list. The container log is the signal, not the deploy output.

The compose header carried two claims that #190 falsified and did not correct: that the PayPal secrets are required "because DEMO_MODE is false below", and that only secrets are interpolated. Both now describe the file as it is.

This is the drift `composeEnvironment.test.ts` exists to prevent, surfacing in the one place no test can reach — the guard keeps the compose file honest about its own intent and cannot see the runbook beside it.

Verified: 278 backend unit tests pass, including the compose guard that parses this file, and a sweep for the stale claims finds none left.

Closes #196

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-08-27 11:35:44 -05:00
co-authored by Claude Opus 5
parent fdb5b7a0be
commit 2fe3aa055f
2 changed files with 29 additions and 12 deletions
+12 -8
View File
@@ -51,12 +51,14 @@
# Portainer's stack variables are substituted into this file; they are not
# handed to the container. A variable set in Portainer with no line here never
# reaches the app, and the failure reads as "I set it and it says it is not
# set". Only secrets are interpolated below, because only secrets have a reason
# not to be in the repository. Everything else is written out, so there is one
# place to look and one thing that can be wrong.
# set". Secrets are interpolated below, because they have a reason not to be in
# the repository — and so is DEMO_MODE, which is not a secret but is the one
# value that gets flipped without a code change, which is a reason of its own
# (#190). Everything else is written out, so there is one place to look and one
# thing that can be wrong.
#
# Required stack environment variables — all secrets, all must be set in
# Portainer for this stack:
# Required stack environment variables. All must be set in Portainer for this
# stack. All are secrets except DEMO_MODE:
#
# DEMO_MODE `true` or `false`, exactly. Whether real payments are
# taken. Not a secret — it is here rather than written
@@ -74,9 +76,11 @@
# See #63. Without it, /api/admin is protected only by
# the proxy — anything reaching the container directly
# can administer the store.
# PAYPAL_CLIENT_ID Live PayPal credentials. Required because DEMO_MODE is
# PAYPAL_CLIENT_SECRET false below; the app refuses to start without them.
# PAYPAL_WEBHOOK_ID
# PAYPAL_CLIENT_ID Live PayPal credentials. Required when DEMO_MODE is
# PAYPAL_CLIENT_SECRET `false`; the app refuses to start without them rather
# PAYPAL_WEBHOOK_ID than pretending it can take payments. Unused while
# DEMO_MODE is `true`, so flipping it to `false` without
# them crash-loops the container.
# BACKUP_PASSPHRASE Optional. Set it and the uploads archives are
# encrypted at rest; leave it empty and they are not.
# See docs/ops/backup-and-restore.md before setting it —
+17 -4
View File
@@ -60,7 +60,9 @@ Everything here is lost when the stack is deleted, and the rollback in step 8 is
**The stack name**, exactly as Portainer shows it. If it is not `redefined-designs`, note that — the new stack must be created with that name, because the stack name becomes the compose project name and reusing QA's would make Compose reconcile the two against each other.
**Every stack environment variable, name and value.** These are all secrets, they belong to the stack, and deleting it discards them. This is the step whose omission is felt hardest: the compose file interpolates `DB_PASSWORD`, `SMTP_USER`, `SMTP_PASSWORD`, `SMTP_FROM`, `ADMIN_GATE_SECRET`, `PAYPAL_CLIENT_ID`, `PAYPAL_CLIENT_SECRET` and `PAYPAL_WEBHOOK_ID`, and an unset one substitutes to an empty string rather than failing. A missing `DB_PASSWORD` cannot authenticate against its own data directory; a missing PayPal secret crash-loops the container, because `DEMO_MODE` is false here and the app refuses to pretend it can take payments. Neither is recoverable from anything in this repository.
**Every stack environment variable, name and value.** They belong to the stack, and deleting it discards them. This is the step whose omission is felt hardest: the compose file interpolates `DEMO_MODE`, `DB_PASSWORD`, `SMTP_USER`, `SMTP_PASSWORD`, `SMTP_FROM`, `ADMIN_GATE_SECRET`, `PAYPAL_CLIENT_ID`, `PAYPAL_CLIENT_SECRET` and `PAYPAL_WEBHOOK_ID`, and an unset one substitutes to an empty string rather than failing. A missing `DB_PASSWORD` cannot authenticate against its own data directory, and none of this is recoverable from anything in this repository.
`DEMO_MODE` is the odd one out and the easiest to miss, because it is the only one that is not a secret. It must be exactly `true` or `false`. There is no default in the compose file, deliberately (#190) — a default would decide whether the shop takes money on the operator's behalf, silently, whichever way it pointed — so an unset `DEMO_MODE` refuses to boot rather than guessing. Production is `true` for now, the interim from #191; setting it to `false` is what restores real payments, and doing that without all three PayPal secrets present crash-loops the container.
Copy them somewhere before you delete anything.
@@ -217,7 +219,16 @@ Migration output must appear *before* `listening on 3000`, and `listening on 300
- `MAIL_ALLOWLIST is not set` — production is the one environment that has to reach real customers.
- `UPLOADS_BASE_URL is not set` — the uploads origin is not configured yet (#103). Expected until an NPM host exists for it.
**What a crash loop looks like, and it is the likeliest outcome of a missed step 2.** A `[config] refusing to start` block, then the whole boot sequence again, repeating:
**What a crash loop looks like, and it is the likeliest outcome of a missed step 2.** A `[config] refusing to start` block, then the whole boot sequence again, repeating.
While production is in the demo interim, the likeliest one is `DEMO_MODE` itself:
```
[config] refusing to start — 1 problem(s) with the environment:
[config] - DEMO_MODE is required and must be exactly 'true' or 'false'.
```
The other form appears once `DEMO_MODE` is `false` and real payments are on:
```
[config] refusing to start — 3 problem(s) with the environment:
@@ -226,9 +237,11 @@ Migration output must appear *before* `listening on 3000`, and `listening on 300
[config] - PAYPAL_WEBHOOK_ID is required when DEMO_MODE=false, because real payments are enabled.
```
That is a secret that did not reach the container, and it is what happens when the stack variables recorded in step 2 were not carried across — they belong to the stack and were discarded with the old one. An unset stack variable substitutes to an empty string rather than failing, so the container receives `PAYPAL_CLIENT_ID=` and refuses it.
Either way it is a stack variable that did not reach the container, and it is what happens when the ones recorded in step 2 were not carried across — they belong to the stack and were discarded with the old one. An unset stack variable substitutes to an empty string rather than failing, so the container receives `PAYPAL_CLIENT_ID=` and refuses it.
Read the list against the compose file to tell a missing variable from a misnamed one. Anything hardcoded there — `PAYPAL_ENV`, `DEMO_MODE`, `UPLOADS_DIR` — cannot be missing, so its absence from the error list proves nothing. What is diagnostic is an *interpolated* variable that is NOT in the list: if `ADMIN_GATE_SECRET` is quiet while the PayPal three complain, substitution is working and those three specifically are unset, rather than something being wrong with the stack.
Compose itself only warns about this — `The "DEMO_MODE" variable is not set. Defaulting to a blank string.` — and deploys anyway. That warning is easy to miss in Portainer's stack UI, and the container then crash-loops under `restart: unless-stopped`: loud in the log, invisible in a glance at the stack list. The container log is the reliable signal, not the deploy output.
Read the list against the compose file to tell a missing variable from a misnamed one. Anything hardcoded there — `PAYPAL_ENV`, `UPLOADS_DIR` — cannot be missing, so its absence from the error list proves nothing. `DEMO_MODE` was in that list and is not any more: since #190 it is interpolated like the secrets, which makes it the first thing to check rather than something to rule out. What is diagnostic is an *interpolated* variable that is NOT in the list: if `ADMIN_GATE_SECRET` is quiet while the PayPal three complain, substitution is working and those three specifically are unset, rather than something being wrong with the stack.
Fix it in the stack's environment, under exactly the names the compose file reads, and redeploy. If the values are gone with the old stack, the client id and secret are in the PayPal developer dashboard under the live app, and the webhook id is on the webhook entry for `PUBLIC_URL/webhooks/paypal` — readable rather than only recreatable.