fix(admin): SMTP over implicit TLS, and rollbacks that survive
Server Deploy / deploy (push) Successful in 2m13s

Two bugs, one symptom: signup created an account and a customer_user but
no verification email ever arrived.

net/smtp.SendMail only speaks STARTTLS. Against a port-465 server, which
expects a TLS handshake immediately, it never delivers. The transport now
wraps the connection before speaking SMTP on 465, exactly as
sitesvc/internal/mail already did — the two are duplicated, so change both
or consolidate into shared/. Also adds Date and Message-ID, whose absence
gets a message scored as spam, and a 15s deadline on the conversation.

The rollbacks ran on the HTTP request's context. A stalled mail server
holds the request until the browser gives up, which cancels that context
and turns both rollbacks into silent no-ops — stranding the exact rows
they exist to remove. They now run detached with their own timeout, and
log when they fail.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
mrhid6
2026-07-26 14:26:34 +01:00
co-authored by Claude Opus 5
parent b209fed4a7
commit aef5811c16
2 changed files with 133 additions and 16 deletions
+23 -3
View File
@@ -5,6 +5,7 @@ import (
"crypto/rand"
"crypto/sha256"
"encoding/hex"
"log"
"net/http"
"strings"
"time"
@@ -61,7 +62,18 @@ func CreateCustomerUser(ctx context.Context, accountID, email, password string)
// worse than no row: it can never be signed in to, and it holds the
// unique index on email, so the customer cannot sign up again with the
// address they just used.
_, _ = db.Admin("customer_users").DeleteOne(ctx, bson.M{"user_id": u.UserID})
//
// Deliberately NOT on ctx. ctx is the HTTP request's, and the most likely
// reason we are here is that the mail server stalled until the browser
// gave up — which cancels ctx and makes this delete a silent no-op,
// stranding exactly the row it exists to remove. That happened in
// production against a port-465 server.
rbCtx, cancel := context.WithTimeout(context.WithoutCancel(ctx), 10*time.Second)
defer cancel()
if _, dErr := db.Admin("customer_users").DeleteOne(rbCtx, bson.M{"user_id": u.UserID}); dErr != nil {
log.Printf("signup: FAILED to roll back customer_user %s (%s) after mail error: %v",
u.UserID, u.Email, dErr)
}
return err
}
return nil
@@ -123,8 +135,16 @@ func HandleSignup(c *gin.Context) {
}
if err := CreateCustomerUser(ctx, acct.AccountID, email, body.Password); err != nil {
// Roll the account back rather than strand one with no owner.
_, _ = db.Admin("accounts").DeleteOne(ctx, bson.M{"account_id": acct.AccountID})
// Roll the account back rather than strand one with no owner. Detached
// from ctx for the same reason as the user rollback above: a stalled mail
// server cancels the request, and a rollback that needs the request to
// still be alive is a rollback that fails exactly when it is needed.
log.Printf("signup: %s failed, rolling back account %s: %v", email, acct.AccountID, err)
rbCtx, cancel := context.WithTimeout(context.WithoutCancel(ctx), 10*time.Second)
defer cancel()
if _, dErr := db.Admin("accounts").DeleteOne(rbCtx, bson.M{"account_id": acct.AccountID}); dErr != nil {
log.Printf("signup: FAILED to roll back account %s: %v", acct.AccountID, dErr)
}
c.JSON(http.StatusInternalServerError, gin.H{"error": "could not send the verification email"})
return
}