BloopAI/vibe-kanban · error · ContainerError

MsgStore missing for execution {} during normalization setup

Error message

MsgStore missing for execution {} during normalization setup

What it means

After creating the execution process record, start_execution looks up an in-memory MsgStore for the new process; it must have been registered by a prior setup step. If it is absent the process record is removed from the map and this error is returned, aborting the execution during normalization setup.

Source

Thrown at crates/services/src/services/container.rs:1323

            )),
            ExecutorActionType::CodingAgentFollowUpRequest(request) => Some((
                request.executor_config.profile_id(),
                request.effective_dir(&workspace_root),
            )),
            ExecutorActionType::ReviewRequest(request) => Some((
                request.executor_config.profile_id(),
                request.effective_dir(&workspace_root),
            )),
            _ => None,
        } {
            let msg_store = match self.get_msg_store_by_id(&execution_process.id).await {
                Some(store) => store,
                None => {
                    self.msg_stores()
                        .write()
                        .await
                        .remove(&execution_process.id);
                    return Err(ContainerError::Other(anyhow!(
                        "MsgStore missing for execution {} during normalization setup",
                        execution_process.id
                    )));
                }
            };
            #[cfg(feature = "qa-mode")]
            {
                let executor = QaMockExecutor;
                let _ = executor.normalize_logs(msg_store, &working_dir);
            }
            #[cfg(not(feature = "qa-mode"))]
            {
                if let Some(executor) =
                    ExecutorConfigs::get_cached().get_coding_agent(&executor_profile_id)
                {
                    let _ = executor.normalize_logs(msg_store, &working_dir);
                } else {
                    tracing::error!(

View on GitHub (pinned to 4deb7eca8f)

Solutions

Retry the execution — if it was a transient race, a new attempt registers a fresh MsgStore.
Check whether another task removes entries from msg_stores and ensure it doesn't target just-created processes.
Ensure single-flight per workspace so concurrent start_execution calls don't clean each other's stores.
If reproducible, capture logs around process creation and file a bug: the store registration path is skipping insertion.

Defensive patterns

Strategy: retry

Try / catch

match start_execution(...).await {
    Err(e) if e.to_string().contains("MsgStore missing") => {
        tokio::time::sleep(Duration::from_millis(500)).await;
        start_execution(...).await
    }
    other => other,
}

Prevention

Avoid concurrent operations on the same execution process (single-flight per process id).
Don't cancel/cleanup a process between creation and its first execution.
Report persistent occurrences — in-memory state should not diverge from the DB.

When it happens

Trigger: start_execution reaching the msg_stores() lookup when the store for execution_process.id was never inserted or was evicted — e.g. internal race, prior cleanup removed it, or a bug skipping store registration.

Common situations: Concurrent cancellation/cleanup deleting the store between process creation and lookup; restarting parts of the service losing in-memory state while DB rows persist; rapid follow-up calls racing with process teardown.

Related errors

AI-assisted analysis of BloopAI/vibe-kanban@4deb7eca8f (2026-08-29). Data as JSON: /api/errors/8d16c718d396e27b. Report an issue: GitHub.