fix(whatsmeow): create the sqlstore container once instead of on every StartClient

[EcoosUP]

The problem

StartClient calls sqlstore.New on every invocation and never closes the previous container. There is no container.Close() anywhere in the file.

Each sqlstore.New opens its own *sql.DB. A *sql.DB that is never closed keeps its TCP connections open for the lifetime of the process — the garbage collector does not reclaim them.

This is harmless while instances stay connected. It becomes fatal when one instance loops, and a device that was logged out from the phone does exactly that:

reconnect → no session → QR → nobody scans → max QR count reached
   → forced logout → Disconnected event → reconnect → ...

Every turn of that loop leaks a whole connection pool.

Measured in production

A 500-connection Postgres, nine instances, one of them logged out from the phone at 21:06:

reconnects in 50 minutes	110
QR codes generated	1135
connections leaked per reconnect	~4.5
time to exhaust the server	44 minutes

At 21:50 the pool was gone, and from that point on every other instance that dropped its websocket failed to come back:

Disconnected detected, restarting instance
Failed to create container: failed to upgrade database:
  failed to check if version table is up to date: pq: sorry, too many clients already

Three unrelated numbers went silent for 14 hours because of one dead instance. All 500 connections were sitting idle when we found them — nothing was using them.

The fix

The DSN is identical for every instance (w.config.PostgresAuthDB, or the same sqlite path), so there was never a reason for a container per instance. This creates it once and shares it.

After the change, same deployment: 10 StartClient calls, 4 connections in use.

Two notes on the implementation

The error is not memoized. A sync.Once would let a database that is briefly unreachable at the first attempt poison the process for its entire lifetime. A Mutex guarding a nil check retries until it succeeds, and only the success is stored.

var err error was removed from the top of StartClient because the new := declares it. No behaviour change.

Testing

Built against main (go build ./..., clean) and running in production on two separate Evolution GO processes.

Happy to split this differently or adjust naming if you prefer.

Summary by Sourcery

Reuse a single whatsmeow SQL store container across client starts to prevent connection exhaustion during reconnect loops.

Bug Fixes:

• Prevent WhatsApp client restarts from leaking database connection pools by reusing a shared SQL store container.

Enhancements:

• Allow shared-container initialization to retry after transient database connection failures instead of permanently caching the initial error.

[sourcery-ai]

Reviewer's Guide

Fixes connection-pool exhaustion by creating the whatsmeow SQL store container once and sharing it across StartClient calls, while allowing transient initialization failures to be retried.

Sequence diagram for shared whatsmeow container initialization

sequenceDiagram
    participant StartClient
    participant sharedContainer
    participant sqlstore
    participant Database

    StartClient->>sharedContainer: sharedContainer(w)
    sharedContainer->>sharedContainer: Lock sharedContainerMu
    alt sharedStore != nil
        sharedContainer-->>StartClient: sharedStore
    else sharedStore is nil
        sharedContainer->>sqlstore: New(context.Background(), driver, dsn, dbLog)
        alt initialization succeeds
            sqlstore->>Database: Open connection pool
            Database-->>sqlstore: Pool ready
            sharedContainer->>sharedContainer: sharedStore = container
            sharedContainer-->>StartClient: container
        else initialization fails
            sqlstore-->>sharedContainer: error
            sharedContainer-->>StartClient: error
            StartClient->>StartClient: Retry on a later call
        end
    end
    sharedContainer->>sharedContainer: Unlock sharedContainerMu

Loading

File-Level Changes

Change	Details	Files
Introduces a lazily initialized, process-wide whatsmeow SQL store container and reuses it across client starts.	Adds a mutex-protected shared container accessor. Preserves PostgreSQL and SQLite DSN selection and debug logging during first initialization. Retries initialization after failures by storing the container only after successful creation. Replaces per-StartClient container creation with the shared accessor and removes the redundant error declaration.	pkg/whatsmeow/service/whatsmeow.go

Possibly linked issues

• #Postgres connection leak: The PR directly fixes the reported leak by sharing one sqlstore container instead of creating one per reconnect.
• #unknown: The PR directly fixes the issue by caching and reusing the sqlstore container instead of creating pools per StartClient call.
• #PostgreSQL connection leak: a new sqlstore pool is created on instance reconnect: The PR directly fixes the issue by sharing one sqlstore container across client starts and reconnects.

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Hey - I've reviewed your changes and they look great!

Sourcery assessment

Needs a human reviewer. The first StartClient call fixes the process-wide database container, so if service instances use different SQLite paths or database credentials, later clients could read or write session data in the wrong database. Reverting stops further misrouting, but records written through the wrong container would remain and require cleanup or recovery.

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook