temporalio 0.0.0 → 0.0.2

data/bridge/sdk-core/core/src/worker/workflow/mod.rs

@@ -0,0 +1,1200 @@
+//! This module and its submodules implement Core's logic for managing workflows (which is the
+//! lion's share of the complexity in Core). See the `ARCHITECTURE.md` file in the repo root for
+//! a diagram of the internals.
+
+mod bridge;
+mod driven_workflow;
+mod history_update;
+mod machines;
+mod managed_run;
+mod run_cache;
+pub(crate) mod wft_poller;
+mod workflow_stream;
+
+pub(crate) use bridge::WorkflowBridge;
+pub(crate) use driven_workflow::{DrivenWorkflow, WorkflowFetcher};
+pub(crate) use history_update::{HistoryPaginator, HistoryUpdate};
+pub(crate) use machines::WFMachinesError;
+#[cfg(test)]
+pub(crate) use managed_run::ManagedWFFunc;
+
+use crate::{
+    abstractions::OwnedMeteredSemPermit,
+    protosext::{legacy_query_failure, ValidPollWFTQResponse, WorkflowActivationExt},
+    telemetry::VecDisplayer,
+    worker::{
+        activities::{ActivitiesFromWFTsHandle, PermittedTqResp},
+        client::{WorkerClient, WorkflowTaskCompletion},
+        workflow::{
+            managed_run::{ManagedRun, WorkflowManager},
+            wft_poller::validate_wft,
+            workflow_stream::{LocalInput, LocalInputs, WFStream},
+        },
+        LocalActRequest, LocalActivityResolution,
+    },
+    MetricsContext,
+};
+use futures::{stream::BoxStream, Stream, StreamExt};
+use std::{
+    collections::HashSet,
+    fmt::{Debug, Display, Formatter},
+    future::Future,
+    ops::DerefMut,
+    result,
+    sync::Arc,
+    time::{Duration, Instant},
+};
+use temporal_sdk_core_api::errors::{CompleteWfError, PollWfError};
+use temporal_sdk_core_protos::{
+    coresdk::{
+        workflow_activation::{
+            remove_from_cache::EvictionReason, workflow_activation_job, QueryWorkflow,
+            WorkflowActivation, WorkflowActivationJob,
+        },
+        workflow_commands::*,
+        workflow_completion,
+        workflow_completion::{
+            workflow_activation_completion, Failure, WorkflowActivationCompletion,
+        },
+    },
+    temporal::api::{
+        command::v1::{command::Attributes, Command as ProtoCommand, Command},
+        common::v1::{Memo, RetryPolicy, SearchAttributes},
+        enums::v1::WorkflowTaskFailedCause,
+        taskqueue::v1::StickyExecutionAttributes,
+        workflowservice::v1::PollActivityTaskQueueResponse,
+    },
+    TaskToken,
+};
+use tokio::{
+    sync::{
+        mpsc::{unbounded_channel, UnboundedSender},
+        oneshot,
+    },
+    task,
+    task::{JoinError, JoinHandle},
+};
+use tokio_stream::wrappers::UnboundedReceiverStream;
+use tokio_util::sync::CancellationToken;
+use tracing::Span;
+
+pub(crate) const LEGACY_QUERY_ID: &str = "legacy_query";
+const MAX_EAGER_ACTIVITY_RESERVATIONS_PER_WORKFLOW_TASK: usize = 3;
+
+type Result<T, E = WFMachinesError> = result::Result<T, E>;
+type BoxedActivationStream = BoxStream<'static, Result<ActivationOrAuto, PollWfError>>;
+
+/// Centralizes all state related to workflows and workflow tasks
+pub(crate) struct Workflows {
+    task_queue: String,
+    local_tx: UnboundedSender<LocalInput>,
+    processing_task: tokio::sync::Mutex<Option<JoinHandle<()>>>,
+    activation_stream: tokio::sync::Mutex<(
+        BoxedActivationStream,
+        // Used to indicate polling may begin
+        Option<oneshot::Sender<()>>,
+    )>,
+    client: Arc<dyn WorkerClient>,
+    /// Will be populated when this worker is using a cache and should complete WFTs with a sticky
+    /// queue.
+    sticky_attrs: Option<StickyExecutionAttributes>,
+    /// If set, can be used to reserve activity task slots for eager-return of new activity tasks.
+    activity_tasks_handle: Option<ActivitiesFromWFTsHandle>,
+}
+
+pub(super) struct WorkflowBasics {
+    pub max_cached_workflows: usize,
+    pub max_outstanding_wfts: usize,
+    pub shutdown_token: CancellationToken,
+    pub metrics: MetricsContext,
+    pub namespace: String,
+    pub task_queue: String,
+    pub ignore_evicts_on_shutdown: bool,
+}
+
+impl Workflows {
+    pub(super) fn new(
+        basics: WorkflowBasics,
+        sticky_attrs: Option<StickyExecutionAttributes>,
+        client: Arc<dyn WorkerClient>,
+        wft_stream: impl Stream<Item = Result<ValidPollWFTQResponse, tonic::Status>> + Send + 'static,
+        local_activity_request_sink: impl Fn(Vec<LocalActRequest>) -> Vec<LocalActivityResolution>
+            + Send
+            + Sync
+            + 'static,
+        activity_tasks_handle: Option<ActivitiesFromWFTsHandle>,
+    ) -> Self {
+        let (local_tx, local_rx) = unbounded_channel();
+        let shutdown_tok = basics.shutdown_token.clone();
+        let task_queue = basics.task_queue.clone();
+        let mut stream = WFStream::build(
+            basics,
+            wft_stream,
+            UnboundedReceiverStream::new(local_rx),
+            client.clone(),
+            local_activity_request_sink,
+        );
+        let (activation_tx, activation_rx) = unbounded_channel();
+        let (start_polling_tx, start_polling_rx) = oneshot::channel();
+        // We must spawn a task to constantly poll the activation stream, because otherwise
+        // activation completions would not cause anything to happen until the next poll.
+        let processing_task = task::spawn(async move {
+            // However, we want to avoid plowing ahead until we've been asked to poll at least once.
+            // This supports activity-only workers.
+            let do_poll = tokio::select! {
+                sp = start_polling_rx => {
+                    sp.is_ok()
+                }
+                _ = shutdown_tok.cancelled() => {
+                    false
+                }
+            };
+            if !do_poll {
+                return;
+            }
+            while let Some(act) = stream.next().await {
+                activation_tx
+                    .send(act)
+                    .expect("Activation processor channel not dropped");
+            }
+        });
+        Self {
+            task_queue,
+            local_tx,
+            processing_task: tokio::sync::Mutex::new(Some(processing_task)),
+            activation_stream: tokio::sync::Mutex::new((
+                UnboundedReceiverStream::new(activation_rx).boxed(),
+                Some(start_polling_tx),
+            )),
+            client,
+            sticky_attrs,
+            activity_tasks_handle,
+        }
+    }
+
+    pub async fn next_workflow_activation(&self) -> Result<WorkflowActivation, PollWfError> {
+        loop {
+            let r = {
+                let mut lock = self.activation_stream.lock().await;
+                let (ref mut stream, ref mut beginner) = lock.deref_mut();
+                if let Some(beginner) = beginner.take() {
+                    let _ = beginner.send(());
+                }
+                stream.next().await.unwrap_or(Err(PollWfError::ShutDown))?
+            };
+            Span::current().record("run_id", r.run_id());
+            match r {
+                ActivationOrAuto::LangActivation(act) | ActivationOrAuto::ReadyForQueries(act) => {
+                    debug!(activation=%act, "Sending activation to lang");
+                    break Ok(act);
+                }
+                ActivationOrAuto::Autocomplete { run_id } => {
+                    self.activation_completed(WorkflowActivationCompletion {
+                        run_id,
+                        status: Some(workflow_completion::Success::from_variants(vec![]).into()),
+                    })
+                    .await?;
+                }
+            }
+        }
+    }
+
+    /// Queue an activation completion for processing, returning a future that will resolve with
+    /// the outcome of that completion. See [ActivationCompletedOutcome].
+    ///
+    /// Returns the most-recently-processed event number for the run
+    pub async fn activation_completed(
+        &self,
+        completion: WorkflowActivationCompletion,
+    ) -> Result<usize, CompleteWfError> {
+        let is_empty_completion = completion.is_empty();
+        let completion = validate_completion(completion)?;
+        let run_id = completion.run_id().to_string();
+        let (tx, rx) = oneshot::channel();
+        let was_sent = self.send_local(WFActCompleteMsg {
+            completion,
+            response_tx: tx,
+        });
+        if !was_sent {
+            if is_empty_completion {
+                // Empty complete which is likely an evict reply, we can just ignore.
+                return Ok(0);
+            }
+            panic!(
+                "A non-empty completion was not processed. Workflow processing may have \
+                 terminated unexpectedly. This is a bug."
+            );
+        }
+
+        let completion_outcome = rx
+            .await
+            .expect("Send half of activation complete response not dropped");
+        let mut wft_from_complete = None;
+        let reported_wft_to_server = match completion_outcome.outcome {
+            ActivationCompleteOutcome::ReportWFTSuccess(report) => match report {
+                ServerCommandsWithWorkflowInfo {
+                    task_token,
+                    action:
+                        ActivationAction::WftComplete {
+                            mut commands,
+                            query_responses,
+                            force_new_wft,
+                        },
+                } => {
+                    let reserved_act_permits =
+                        self.reserve_activity_slots_for_outgoing_commands(commands.as_mut_slice());
+                    debug!(commands=%commands.display(), query_responses=%query_responses.display(),
+                           force_new_wft, "Sending responses to server");
+                    let mut completion = WorkflowTaskCompletion {
+                        task_token,
+                        commands,
+                        query_responses,
+                        sticky_attributes: None,
+                        return_new_workflow_task: true,
+                        force_create_new_workflow_task: force_new_wft,
+                    };
+                    let sticky_attrs = self.sticky_attrs.clone();
+                    // Do not return new WFT if we would not cache, because returned new WFTs are
+                    // always partial.
+                    if sticky_attrs.is_none() {
+                        completion.return_new_workflow_task = false;
+                    }
+                    completion.sticky_attributes = sticky_attrs;
+
+                    self.handle_wft_reporting_errs(&run_id, || async {
+                        let maybe_wft = self.client.complete_workflow_task(completion).await?;
+                        if let Some(wft) = maybe_wft.workflow_task {
+                            wft_from_complete = Some(validate_wft(wft)?);
+                        }
+                        self.handle_eager_activities(
+                            reserved_act_permits,
+                            maybe_wft.activity_tasks,
+                        );
+                        Ok(())
+                    })
+                    .await;
+                    true
+                }
+                ServerCommandsWithWorkflowInfo {
+                    task_token,
+                    action: ActivationAction::RespondLegacyQuery { result },
+                } => {
+                    self.respond_legacy_query(task_token, *result).await;
+                    true
+                }
+            },
+            ActivationCompleteOutcome::ReportWFTFail(outcome) => match outcome {
+                FailedActivationWFTReport::Report(tt, cause, failure) => {
+                    warn!(run_id=%run_id, failure=?failure, "Failing workflow task");
+                    self.handle_wft_reporting_errs(&run_id, || async {
+                        self.client
+                            .fail_workflow_task(tt, cause, failure.failure.map(Into::into))
+                            .await
+                    })
+                    .await;
+                    true
+                }
+                FailedActivationWFTReport::ReportLegacyQueryFailure(task_token, failure) => {
+                    warn!(run_id=%run_id, failure=?failure, "Failing legacy query request");
+                    self.respond_legacy_query(task_token, legacy_query_failure(failure))
+                        .await;
+                    true
+                }
+            },
+            ActivationCompleteOutcome::DoNothing => false,
+        };
+
+        self.post_activation(PostActivationMsg {
+            run_id,
+            reported_wft_to_server,
+            wft_from_complete,
+        });
+
+        Ok(completion_outcome.most_recently_processed_event)
+    }
+
+    /// Tell workflow that a local activity has finished with the provided result
+    pub fn notify_of_local_result(&self, run_id: impl Into<String>, resolved: LocalResolution) {
+        self.send_local(LocalResolutionMsg {
+            run_id: run_id.into(),
+            res: resolved,
+        });
+    }
+
+    /// Request eviction of a workflow
+    pub fn request_eviction(
+        &self,
+        run_id: impl Into<String>,
+        message: impl Into<String>,
+        reason: EvictionReason,
+    ) {
+        self.send_local(RequestEvictMsg {
+            run_id: run_id.into(),
+            message: message.into(),
+            reason,
+        });
+    }
+
+    /// Query the state of workflow management. Can return `None` if workflow state is shut down.
+    pub fn get_state_info(&self) -> impl Future<Output = Option<WorkflowStateInfo>> {
+        let (tx, rx) = oneshot::channel();
+        self.send_local(GetStateInfoMsg { response_tx: tx });
+        async move { rx.await.ok() }
+    }
+
+    pub async fn shutdown(&self) -> Result<(), JoinError> {
+        let maybe_jh = self.processing_task.lock().await.take();
+        if let Some(jh) = maybe_jh {
+            // This acts as a final wake up in case the stream is still alive and wouldn't otherwise
+            // receive another message. It allows it to shut itself down.
+            let _ = self.get_state_info();
+            jh.await
+        } else {
+            Ok(())
+        }
+    }
+
+    /// Must be called after every activation completion has finished
+    fn post_activation(&self, msg: PostActivationMsg) {
+        self.send_local(msg);
+    }
+
+    /// Handle server errors from either completing or failing a workflow task. Un-handleable errors
+    /// trigger a workflow eviction and are logged.
+    async fn handle_wft_reporting_errs<T, Fut>(&self, run_id: &str, completer: impl FnOnce() -> Fut)
+    where
+        Fut: Future<Output = Result<T, tonic::Status>>,
+    {
+        let mut should_evict = None;
+        if let Err(err) = completer().await {
+            match err.code() {
+                // Silence unhandled command errors since the lang SDK cannot do anything
+                // about them besides poll again, which it will do anyway.
+                tonic::Code::InvalidArgument if err.message() == "UnhandledCommand" => {
+                    debug!(error = %err, run_id, "Unhandled command response when completing");
+                    should_evict = Some(EvictionReason::UnhandledCommand);
+                }
+                tonic::Code::NotFound => {
+                    warn!(error = %err, run_id, "Task not found when completing");
+                    should_evict = Some(EvictionReason::TaskNotFound);
+                }
+                _ => {
+                    warn!(error= %err, "Network error while completing workflow activation");
+                    should_evict = Some(EvictionReason::Fatal);
+                }
+            }
+        }
+        if let Some(reason) = should_evict {
+            self.request_eviction(run_id, "Error reporting WFT to server", reason);
+        }
+    }
+
+    /// Sends a message to the workflow processing stream. Returns true if the message was sent
+    /// successfully.
+    fn send_local(&self, msg: impl Into<LocalInputs>) -> bool {
+        let msg = msg.into();
+        let print_err = !matches!(msg, LocalInputs::GetStateInfo(_));
+        if let Err(e) = self.local_tx.send(LocalInput {
+            input: msg,
+            span: Span::current(),
+        }) {
+            if print_err {
+                warn!(
+                    "Tried to interact with workflow state after it shut down. This may be benign \
+                     when processing evictions during shutdown. When sending {:?}",
+                    e.0.input
+                )
+            }
+            false
+        } else {
+            true
+        }
+    }
+
+    /// Process eagerly returned activities from WFT completion
+    fn handle_eager_activities(
+        &self,
+        reserved_act_permits: Vec<OwnedMeteredSemPermit>,
+        eager_acts: Vec<PollActivityTaskQueueResponse>,
+    ) {
+        if let Some(at_handle) = self.activity_tasks_handle.as_ref() {
+            let excess_reserved = reserved_act_permits.len().saturating_sub(eager_acts.len());
+            if excess_reserved > 0 {
+                debug!(
+                    "Server returned {excess_reserved} fewer activities for \
+                     eager execution than we requested"
+                );
+            } else if eager_acts.len() > reserved_act_permits.len() {
+                // If we somehow got more activities from server than we asked for, server did
+                // something wrong.
+                error!(
+                    "Server sent more activities for eager execution than we requested! They will \
+                     be dropped and eventually time out. Please report this, as it is a server bug."
+                )
+            }
+            let with_permits = reserved_act_permits
+                .into_iter()
+                .zip(eager_acts.into_iter())
+                .map(|(permit, resp)| PermittedTqResp { permit, resp });
+            if with_permits.len() > 0 {
+                debug!(
+                    "Adding {} activity tasks received from WFT complete",
+                    with_permits.len()
+                );
+                at_handle.add_tasks(with_permits);
+            }
+        } else if !eager_acts.is_empty() {
+            panic!(
+                "Requested eager activity execution but this worker has no activity task \
+                 manager! This is an internal bug, Core should not have asked for tasks."
+            )
+        }
+    }
+
+    /// Attempt to reserve activity slots for activities we could eagerly execute on
+    /// this worker.
+    ///
+    /// Returns the number of activity slots that were reserved
+    fn reserve_activity_slots_for_outgoing_commands(
+        &self,
+        commands: &mut [Command],
+    ) -> Vec<OwnedMeteredSemPermit> {
+        let mut reserved = vec![];
+        for cmd in commands {
+            if let Some(Attributes::ScheduleActivityTaskCommandAttributes(attrs)) =
+                cmd.attributes.as_mut()
+            {
+                // If request_eager_execution was already false, that means lang explicitly
+                // told us it didn't want to eagerly execute for some reason. So, we only
+                // ever turn *off* eager execution if a slot is not available or the activity
+                // is scheduled on a different task queue.
+                if attrs.request_eager_execution {
+                    let same_task_queue = attrs
+                        .task_queue
+                        .as_ref()
+                        .map(|q| q.name == self.task_queue)
+                        .unwrap_or_default();
+                    if same_task_queue
+                        && reserved.len() < MAX_EAGER_ACTIVITY_RESERVATIONS_PER_WORKFLOW_TASK
+                    {
+                        if let Some(p) = self
+                            .activity_tasks_handle
+                            .as_ref()
+                            .and_then(|h| h.reserve_slot())
+                        {
+                            reserved.push(p);
+                        } else {
+                            attrs.request_eager_execution = false;
+                        }
+                    } else {
+                        attrs.request_eager_execution = false;
+                    }
+                }
+            }
+        }
+        reserved
+    }
+
+    /// Wraps responding to legacy queries. Handles ignore-able failures.
+    async fn respond_legacy_query(&self, tt: TaskToken, res: QueryResult) {
+        match self.client.respond_legacy_query(tt, res).await {
+            Ok(_) => {}
+            Err(e) if e.code() == tonic::Code::NotFound => {
+                warn!(error=?e, "Query not found when attempting to respond to it");
+            }
+            Err(e) => {
+                warn!(error= %e, "Network error while responding to legacy query");
+            }
+        }
+    }
+}
+
+/// Manages access to a specific workflow run, and contains various bookkeeping information that the
+/// [WFStream] may need to access quickly.
+#[derive(derive_more::DebugCustom)]
+#[debug(
+    fmt = "ManagedRunHandle {{ wft: {:?}, activation: {:?}, buffered_resp: {:?} \
+           have_seen_terminal_event: {}, most_recently_processed_event: {}, more_pending_work: {}, \
+           trying_to_evict: {}, last_action_acked: {} }}",
+    wft,
+    activation,
+    buffered_resp,
+    have_seen_terminal_event,
+    most_recently_processed_event_number,
+    more_pending_work,
+    "trying_to_evict.is_some()",
+    last_action_acked
+)]
+struct ManagedRunHandle {
+    /// If set, the WFT this run is currently/will be processing.
+    wft: Option<OutstandingTask>,
+    /// An outstanding activation to lang
+    activation: Option<OutstandingActivation>,
+    /// If set, it indicates there is a buffered poll response from the server that applies to this
+    /// run. This can happen when lang takes too long to complete a task and the task times out, for
+    /// example. Upon next completion, the buffered response will be removed and can be made ready
+    /// to be returned from polling
+    buffered_resp: Option<PermittedWFT>,
+    /// True if this machine has seen an event which ends the execution
+    have_seen_terminal_event: bool,
+    /// The most recently processed event id this machine has seen. 0 means it has seen nothing.
+    most_recently_processed_event_number: usize,
+    /// Is set true when the machines indicate that there is additional known work to be processed
+    more_pending_work: bool,
+    /// Is set if an eviction has been requested for this run
+    trying_to_evict: Option<RequestEvictMsg>,
+    /// Set to true if the last action we tried to take to this run has been processed (ie: the
+    /// [RunUpdateResponse] for it has been seen.
+    last_action_acked: bool,
+    /// For sending work to the machines
+    run_actions_tx: UnboundedSender<RunAction>,
+    /// Handle to the task where the actual machines live
+    handle: JoinHandle<()>,
+
+    /// We track if we have recorded useful debugging values onto a certain span yet, to overcome
+    /// duplicating field values. Remove this once https://github.com/tokio-rs/tracing/issues/2334
+    /// is fixed.
+    recorded_span_ids: HashSet<tracing::Id>,
+    metrics: MetricsContext,
+}
+impl ManagedRunHandle {
+    fn new(
+        wfm: WorkflowManager,
+        activations_tx: UnboundedSender<RunUpdateResponse>,
+        local_activity_request_sink: LocalActivityRequestSink,
+        metrics: MetricsContext,
+    ) -> Self {
+        let (run_actions_tx, run_actions_rx) = unbounded_channel();
+        let managed = ManagedRun::new(wfm, activations_tx, local_activity_request_sink);
+        let handle = tokio::task::spawn(managed.run(run_actions_rx));
+        Self {
+            wft: None,
+            activation: None,
+            buffered_resp: None,
+            have_seen_terminal_event: false,
+            most_recently_processed_event_number: 0,
+            more_pending_work: false,
+            trying_to_evict: None,
+            last_action_acked: true,
+            run_actions_tx,
+            handle,
+            recorded_span_ids: Default::default(),
+            metrics,
+        }
+    }
+
+    fn incoming_wft(&mut self, wft: NewIncomingWFT) {
+        if self.wft.is_some() {
+            error!("Trying to send a new WFT for a run which already has one!");
+        }
+        self.send_run_action(RunActions::NewIncomingWFT(wft));
+    }
+    fn check_more_activations(&mut self) {
+        // No point in checking for more activations if we have not acked the last update, or
+        // if there's already an outstanding activation.
+        if self.last_action_acked && self.activation.is_none() {
+            self.send_run_action(RunActions::CheckMoreWork {
+                want_to_evict: self.trying_to_evict.clone(),
+                has_pending_queries: self
+                    .wft
+                    .as_ref()
+                    .map(|wft| !wft.pending_queries.is_empty())
+                    .unwrap_or_default(),
+                has_wft: self.wft.is_some(),
+            });
+        }
+    }
+    fn send_completion(&mut self, c: RunActivationCompletion) {
+        self.send_run_action(RunActions::ActivationCompletion(c));
+    }
+    fn send_local_resolution(&mut self, r: LocalResolution) {
+        self.send_run_action(RunActions::LocalResolution(r));
+    }
+
+    fn insert_outstanding_activation(&mut self, act: &ActivationOrAuto) {
+        let act_type = match &act {
+            ActivationOrAuto::LangActivation(act) | ActivationOrAuto::ReadyForQueries(act) => {
+                if act.is_legacy_query() {
+                    OutstandingActivation::LegacyQuery
+                } else {
+                    OutstandingActivation::Normal {
+                        contains_eviction: act.eviction_index().is_some(),
+                        num_jobs: act.jobs.len(),
+                    }
+                }
+            }
+            ActivationOrAuto::Autocomplete { .. } => OutstandingActivation::Autocomplete,
+        };
+        if let Some(old_act) = self.activation {
+            // This is a panic because we have screwed up core logic if this is violated. It must be
+            // upheld.
+            panic!(
+                "Attempted to insert a new outstanding activation {:?}, but there already was \
+                 one outstanding: {:?}",
+                act, old_act
+            );
+        }
+        self.activation = Some(act_type);
+    }
+
+    fn send_run_action(&mut self, action: RunActions) {
+        self.last_action_acked = false;
+        self.run_actions_tx
+            .send(RunAction {
+                action,
+                trace_span: Span::current(),
+            })
+            .expect("Receive half of run actions not dropped");
+    }
+
+    /// Returns true if the managed run has any form of pending work
+    /// If `ignore_evicts` is true, pending evictions do not count as pending work.
+    /// If `ignore_buffered` is true, buffered workflow tasks do not count as pending work.
+    fn has_any_pending_work(&self, ignore_evicts: bool, ignore_buffered: bool) -> bool {
+        let evict_work = if ignore_evicts {
+            false
+        } else {
+            self.trying_to_evict.is_some()
+        };
+        let act_work = if ignore_evicts {
+            if let Some(ref act) = self.activation {
+                !act.has_only_eviction()
+            } else {
+                false
+            }
+        } else {
+            self.activation.is_some()
+        };
+        let buffered = if ignore_buffered {
+            false
+        } else {
+            self.buffered_resp.is_some()
+        };
+        self.wft.is_some()
+            || buffered
+            || !self.last_action_acked
+            || self.more_pending_work
+            || act_work
+            || evict_work
+    }
+
+    /// Returns true if the handle is currently processing a WFT which contains a legacy query.
+    fn pending_work_is_legacy_query(&self) -> bool {
+        // Either we know because there is a pending legacy query, or it's already been drained and
+        // sent as an activation.
+        matches!(self.activation, Some(OutstandingActivation::LegacyQuery))
+            || self
+                .wft
+                .as_ref()
+                .map(|t| t.has_pending_legacy_query())
+                .unwrap_or_default()
+    }
+}
+
+#[derive(Debug, derive_more::Display)]
+enum ActivationOrAuto {
+    LangActivation(WorkflowActivation),
+    /// This type should only be filled with an empty activation which is ready to have queries
+    /// inserted into the joblist
+    ReadyForQueries(WorkflowActivation),
+    Autocomplete {
+        run_id: String,
+    },
+}
+impl ActivationOrAuto {
+    pub fn run_id(&self) -> &str {
+        match self {
+            ActivationOrAuto::LangActivation(act) => &act.run_id,
+            ActivationOrAuto::Autocomplete { run_id, .. } => run_id,
+            ActivationOrAuto::ReadyForQueries(act) => &act.run_id,
+        }
+    }
+}
+
+#[derive(derive_more::DebugCustom)]
+#[debug(fmt = "PermittedWft {{ {:?} }}", wft)]
+pub(crate) struct PermittedWFT {
+    wft: ValidPollWFTQResponse,
+    permit: OwnedMeteredSemPermit,
+}
+
+#[derive(Debug)]
+pub(crate) struct OutstandingTask {
+    pub info: WorkflowTaskInfo,
+    pub hit_cache: bool,
+    /// Set if the outstanding task has quer(ies) which must be fulfilled upon finishing replay
+    pub pending_queries: Vec<QueryWorkflow>,
+    pub start_time: Instant,
+    /// The WFT permit owned by this task, ensures we don't exceed max concurrent WFT, and makes
+    /// sure the permit is automatically freed when we delete the task.
+    pub permit: OwnedMeteredSemPermit,
+}
+
+impl OutstandingTask {
+    pub fn has_pending_legacy_query(&self) -> bool {
+        self.pending_queries
+            .iter()
+            .any(|q| q.query_id == LEGACY_QUERY_ID)
+    }
+}
+
+#[derive(Copy, Clone, Debug)]
+pub(crate) enum OutstandingActivation {
+    /// A normal activation with a joblist
+    Normal {
+        /// True if there is an eviction in the joblist
+        contains_eviction: bool,
+        /// Number of jobs in the activation
+        num_jobs: usize,
+    },
+    /// An activation for a legacy query
+    LegacyQuery,
+    /// A fake activation which is never sent to lang, but used internally
+    Autocomplete,
+}
+
+impl OutstandingActivation {
+    pub(crate) const fn has_only_eviction(self) -> bool {
+        matches!(
+            self,
+            OutstandingActivation::Normal {
+                contains_eviction: true,
+                num_jobs: nj
+            }
+        if nj == 1)
+    }
+    pub(crate) const fn has_eviction(self) -> bool {
+        matches!(
+            self,
+            OutstandingActivation::Normal {
+                contains_eviction: true,
+                ..
+            }
+        )
+    }
+}
+
+/// Contains important information about a given workflow task that we need to memorize while
+/// lang handles it.
+#[derive(Clone, Debug)]
+pub struct WorkflowTaskInfo {
+    pub task_token: TaskToken,
+    pub attempt: u32,
+    /// Exists to allow easy tagging of spans with workflow ids. Is duplicative of info inside the
+    /// run machines themselves, but that can't be accessed easily. Would be nice to somehow have a
+    /// shared repository, or refcounts, or whatever, for strings like these that get duped all
+    /// sorts of places.
+    pub wf_id: String,
+}
+
+#[derive(Debug)]
+pub enum FailedActivationWFTReport {
+    Report(TaskToken, WorkflowTaskFailedCause, Failure),
+    ReportLegacyQueryFailure(TaskToken, Failure),
+}
+
+#[derive(Debug)]
+pub(crate) struct ServerCommandsWithWorkflowInfo {
+    pub task_token: TaskToken,
+    pub action: ActivationAction,
+}
+
+#[derive(Debug)]
+pub(crate) enum ActivationAction {
+    /// We should respond that the workflow task is complete
+    WftComplete {
+        commands: Vec<ProtoCommand>,
+        query_responses: Vec<QueryResult>,
+        force_new_wft: bool,
+    },
+    /// We should respond to a legacy query request
+    RespondLegacyQuery { result: Box<QueryResult> },
+}
+
+#[derive(Debug, Eq, PartialEq, Hash)]
+pub(crate) enum EvictionRequestResult {
+    EvictionRequested(Option<u32>),
+    NotFound,
+    EvictionAlreadyRequested(Option<u32>),
+}
+
+#[derive(Debug)]
+#[allow(dead_code)] // Not always used in non-test
+pub(crate) struct WorkflowStateInfo {
+    pub cached_workflows: usize,
+    pub outstanding_wft: usize,
+    pub available_wft_permits: usize,
+}
+
+#[derive(Debug)]
+struct WFActCompleteMsg {
+    completion: ValidatedCompletion,
+    response_tx: oneshot::Sender<ActivationCompleteResult>,
+}
+#[derive(Debug)]
+struct LocalResolutionMsg {
+    run_id: String,
+    res: LocalResolution,
+}
+#[derive(Debug)]
+struct PostActivationMsg {
+    run_id: String,
+    reported_wft_to_server: bool,
+    wft_from_complete: Option<ValidPollWFTQResponse>,
+}
+#[derive(Debug, Clone)]
+struct RequestEvictMsg {
+    run_id: String,
+    message: String,
+    reason: EvictionReason,
+}
+#[derive(Debug)]
+struct GetStateInfoMsg {
+    response_tx: oneshot::Sender<WorkflowStateInfo>,
+}
+
+/// Each activation completion produces one of these
+#[derive(Debug)]
+struct ActivationCompleteResult {
+    most_recently_processed_event: usize,
+    outcome: ActivationCompleteOutcome,
+}
+/// What needs to be done after calling [Workflows::activation_completed]
+#[derive(Debug)]
+#[allow(clippy::large_enum_variant)]
+enum ActivationCompleteOutcome {
+    /// The WFT must be reported as successful to the server using the contained information.
+    ReportWFTSuccess(ServerCommandsWithWorkflowInfo),
+    /// The WFT must be reported as failed to the server using the contained information.
+    ReportWFTFail(FailedActivationWFTReport),
+    /// There's nothing to do right now. EX: The workflow needs to keep replaying.
+    DoNothing,
+}
+#[derive(Debug)]
+struct FulfillableActivationComplete {
+    result: ActivationCompleteResult,
+    resp_chan: oneshot::Sender<ActivationCompleteResult>,
+}
+impl FulfillableActivationComplete {
+    fn fulfill(self) {
+        let _ = self.resp_chan.send(self.result);
+    }
+}
+
+fn validate_completion(
+    completion: WorkflowActivationCompletion,
+) -> Result<ValidatedCompletion, CompleteWfError> {
+    match completion.status {
+        Some(workflow_activation_completion::Status::Successful(success)) => {
+            // Convert to wf commands
+            let commands = success
+                .commands
+                .into_iter()
+                .map(|c| c.try_into())
+                .collect::<Result<Vec<_>, EmptyWorkflowCommandErr>>()
+                .map_err(|_| CompleteWfError::MalformedWorkflowCompletion {
+                    reason: "At least one workflow command in the completion contained \
+                                 an empty variant"
+                        .to_owned(),
+                    run_id: completion.run_id.clone(),
+                })?;
+
+            if commands.len() > 1
+                && commands.iter().any(
+                    |c| matches!(c, WFCommand::QueryResponse(q) if q.query_id == LEGACY_QUERY_ID),
+                )
+            {
+                return Err(CompleteWfError::MalformedWorkflowCompletion {
+                    reason: format!(
+                        "Workflow completion had a legacy query response along with other \
+                         commands. This is not allowed and constitutes an error in the \
+                         lang SDK. Commands: {:?}",
+                        commands
+                    ),
+                    run_id: completion.run_id,
+                });
+            }
+
+            Ok(ValidatedCompletion::Success {
+                run_id: completion.run_id,
+                commands,
+            })
+        }
+        Some(workflow_activation_completion::Status::Failed(failure)) => {
+            Ok(ValidatedCompletion::Fail {
+                run_id: completion.run_id,
+                failure,
+            })
+        }
+        None => Err(CompleteWfError::MalformedWorkflowCompletion {
+            reason: "Workflow completion had empty status field".to_owned(),
+            run_id: completion.run_id,
+        }),
+    }
+}
+
+#[derive(Debug)]
+#[allow(clippy::large_enum_variant)]
+enum ValidatedCompletion {
+    Success {
+        run_id: String,
+        commands: Vec<WFCommand>,
+    },
+    Fail {
+        run_id: String,
+        failure: Failure,
+    },
+}
+
+impl ValidatedCompletion {
+    pub fn run_id(&self) -> &str {
+        match self {
+            ValidatedCompletion::Success { run_id, .. } => run_id,
+            ValidatedCompletion::Fail { run_id, .. } => run_id,
+        }
+    }
+}
+
+/// Input to run tasks, sent to [ManagedRun]s via [ManagedRunHandle]s
+#[derive(Debug)]
+struct RunAction {
+    action: RunActions,
+    trace_span: Span,
+}
+#[derive(Debug)]
+#[allow(clippy::large_enum_variant)]
+enum RunActions {
+    NewIncomingWFT(NewIncomingWFT),
+    ActivationCompletion(RunActivationCompletion),
+    CheckMoreWork {
+        want_to_evict: Option<RequestEvictMsg>,
+        has_pending_queries: bool,
+        has_wft: bool,
+    },
+    LocalResolution(LocalResolution),
+    HeartbeatTimeout,
+}
+#[derive(Debug)]
+struct NewIncomingWFT {
+    /// This field is only populated if the machines already exist. Otherwise the machines
+    /// are instantiated with the workflow history.
+    history_update: Option<HistoryUpdate>,
+    /// Wft start time
+    start_time: Instant,
+}
+#[derive(Debug)]
+struct RunActivationCompletion {
+    task_token: TaskToken,
+    start_time: Instant,
+    commands: Vec<WFCommand>,
+    activation_was_eviction: bool,
+    activation_was_only_eviction: bool,
+    has_pending_query: bool,
+    query_responses: Vec<QueryResult>,
+    /// Used to notify the worker when the completion is done processing and the completion can
+    /// unblock. Must always be `Some` when initialized.
+    resp_chan: Option<oneshot::Sender<ActivationCompleteResult>>,
+}
+
+/// A response from a [ManagedRun] held by a [ManagedRunHandle]
+#[derive(Debug)]
+struct RunUpdateResponse {
+    kind: RunUpdateResponseKind,
+    span: Span,
+}
+#[derive(Debug, derive_more::Display)]
+#[allow(clippy::large_enum_variant)]
+enum RunUpdateResponseKind {
+    Good(GoodRunUpdate),
+    Fail(FailRunUpdate),
+}
+impl RunUpdateResponseKind {
+    pub(crate) fn run_id(&self) -> &str {
+        match self {
+            RunUpdateResponseKind::Good(g) => &g.run_id,
+            RunUpdateResponseKind::Fail(f) => &f.run_id,
+        }
+    }
+}
+
+#[derive(Debug)]
+struct GoodRunUpdate {
+    run_id: String,
+    outgoing_activation: Option<ActivationOrAuto>,
+    fulfillable_complete: Option<FulfillableActivationComplete>,
+    have_seen_terminal_event: bool,
+    /// Is true if there are more jobs that need to be sent to lang
+    more_pending_work: bool,
+    most_recently_processed_event_number: usize,
+    /// Is true if this update was in response to a new WFT
+    in_response_to_wft: bool,
+}
+impl Display for GoodRunUpdate {
+    fn fmt(&self, f: &mut Formatter<'_>) -> std::fmt::Result {
+        write!(
+            f,
+            "GoodRunUpdate(run_id: {}, outgoing_activation: {}, more_pending_work: {})",
+            self.run_id,
+            if let Some(og) = self.outgoing_activation.as_ref() {
+                format!("{}", og)
+            } else {
+                "None".to_string()
+            },
+            self.more_pending_work
+        )
+    }
+}
+#[derive(Debug)]
+pub(crate) struct FailRunUpdate {
+    run_id: String,
+    err: WFMachinesError,
+    /// This is populated if the run update failed while processing a completion - and thus we
+    /// must respond down it when handling the failure.
+    completion_resp: Option<oneshot::Sender<ActivationCompleteResult>>,
+}
+impl Display for FailRunUpdate {
+    fn fmt(&self, f: &mut Formatter<'_>) -> std::fmt::Result {
+        write!(
+            f,
+            "FailRunUpdate(run_id: {}, error: {:?})",
+            self.run_id, self.err
+        )
+    }
+}
+#[derive(Debug)]
+pub struct OutgoingServerCommands {
+    pub commands: Vec<ProtoCommand>,
+    pub replaying: bool,
+}
+
+#[derive(Debug)]
+pub(crate) enum LocalResolution {
+    LocalActivity(LocalActivityResolution),
+}
+
+#[derive(thiserror::Error, Debug, derive_more::From)]
+#[error("Lang provided workflow command with empty variant")]
+pub struct EmptyWorkflowCommandErr;
+
+/// [DrivenWorkflow]s respond with these when called, to indicate what they want to do next.
+/// EX: Create a new timer, complete the workflow, etc.
+#[derive(Debug, derive_more::From, derive_more::Display)]
+#[allow(clippy::large_enum_variant)]
+pub enum WFCommand {
+    /// Returned when we need to wait for the lang sdk to send us something
+    NoCommandsFromLang,
+    AddActivity(ScheduleActivity),
+    AddLocalActivity(ScheduleLocalActivity),
+    RequestCancelActivity(RequestCancelActivity),
+    RequestCancelLocalActivity(RequestCancelLocalActivity),
+    AddTimer(StartTimer),
+    CancelTimer(CancelTimer),
+    CompleteWorkflow(CompleteWorkflowExecution),
+    FailWorkflow(FailWorkflowExecution),
+    QueryResponse(QueryResult),
+    ContinueAsNew(ContinueAsNewWorkflowExecution),
+    CancelWorkflow(CancelWorkflowExecution),
+    SetPatchMarker(SetPatchMarker),
+    AddChildWorkflow(StartChildWorkflowExecution),
+    CancelChild(CancelChildWorkflowExecution),
+    RequestCancelExternalWorkflow(RequestCancelExternalWorkflowExecution),
+    SignalExternalWorkflow(SignalExternalWorkflowExecution),
+    CancelSignalWorkflow(CancelSignalWorkflow),
+    UpsertSearchAttributes(UpsertWorkflowSearchAttributes),
+    ModifyWorkflowProperties(ModifyWorkflowProperties),
+}
+
+impl TryFrom<WorkflowCommand> for WFCommand {
+    type Error = EmptyWorkflowCommandErr;
+
+    fn try_from(c: WorkflowCommand) -> result::Result<Self, Self::Error> {
+        match c.variant.ok_or(EmptyWorkflowCommandErr)? {
+            workflow_command::Variant::StartTimer(s) => Ok(Self::AddTimer(s)),
+            workflow_command::Variant::CancelTimer(s) => Ok(Self::CancelTimer(s)),
+            workflow_command::Variant::ScheduleActivity(s) => Ok(Self::AddActivity(s)),
+            workflow_command::Variant::RequestCancelActivity(s) => {
+                Ok(Self::RequestCancelActivity(s))
+            }
+            workflow_command::Variant::CompleteWorkflowExecution(c) => {
+                Ok(Self::CompleteWorkflow(c))
+            }
+            workflow_command::Variant::FailWorkflowExecution(s) => Ok(Self::FailWorkflow(s)),
+            workflow_command::Variant::RespondToQuery(s) => Ok(Self::QueryResponse(s)),
+            workflow_command::Variant::ContinueAsNewWorkflowExecution(s) => {
+                Ok(Self::ContinueAsNew(s))
+            }
+            workflow_command::Variant::CancelWorkflowExecution(s) => Ok(Self::CancelWorkflow(s)),
+            workflow_command::Variant::SetPatchMarker(s) => Ok(Self::SetPatchMarker(s)),
+            workflow_command::Variant::StartChildWorkflowExecution(s) => {
+                Ok(Self::AddChildWorkflow(s))
+            }
+            workflow_command::Variant::RequestCancelExternalWorkflowExecution(s) => {
+                Ok(Self::RequestCancelExternalWorkflow(s))
+            }
+            workflow_command::Variant::SignalExternalWorkflowExecution(s) => {
+                Ok(Self::SignalExternalWorkflow(s))
+            }
+            workflow_command::Variant::CancelSignalWorkflow(s) => Ok(Self::CancelSignalWorkflow(s)),
+            workflow_command::Variant::CancelChildWorkflowExecution(s) => Ok(Self::CancelChild(s)),
+            workflow_command::Variant::ScheduleLocalActivity(s) => Ok(Self::AddLocalActivity(s)),
+            workflow_command::Variant::RequestCancelLocalActivity(s) => {
+                Ok(Self::RequestCancelLocalActivity(s))
+            }
+            workflow_command::Variant::UpsertWorkflowSearchAttributes(s) => {
+                Ok(Self::UpsertSearchAttributes(s))
+            }
+            workflow_command::Variant::ModifyWorkflowProperties(s) => {
+                Ok(Self::ModifyWorkflowProperties(s))
+            }
+        }
+    }
+}
+
+#[derive(Debug, PartialEq, Eq, Hash, Clone, Copy)]
+enum CommandID {
+    Timer(u32),
+    Activity(u32),
+    LocalActivity(u32),
+    ChildWorkflowStart(u32),
+    SignalExternal(u32),
+    CancelExternal(u32),
+}
+
+/// Details remembered from the workflow execution started event that we may need to recall later.
+/// Is a subset of `WorkflowExecutionStartedEventAttributes`, but avoids holding on to huge fields.
+#[derive(Debug, Clone)]
+pub struct WorkflowStartedInfo {
+    workflow_task_timeout: Option<Duration>,
+    workflow_execution_timeout: Option<Duration>,
+    memo: Option<Memo>,
+    search_attrs: Option<SearchAttributes>,
+    retry_policy: Option<RetryPolicy>,
+}
+
+type LocalActivityRequestSink =
+    Arc<dyn Fn(Vec<LocalActRequest>) -> Vec<LocalActivityResolution> + Send + Sync>;
+
+/// Wraps outgoing activation job protos with some internal details core might care about
+#[derive(Debug, derive_more::Display)]
+#[display(fmt = "{}", variant)]
+struct OutgoingJob {
+    variant: workflow_activation_job::Variant,
+    /// Since LA resolutions are not distinguished from non-LA resolutions as far as lang is
+    /// concerned, but core cares about that sometimes, attach that info here.
+    is_la_resolution: bool,
+}
+impl<WA: Into<workflow_activation_job::Variant>> From<WA> for OutgoingJob {
+    fn from(wa: WA) -> Self {
+        Self {
+            variant: wa.into(),
+            is_la_resolution: false,
+        }
+    }
+}
+impl From<OutgoingJob> for WorkflowActivationJob {
+    fn from(og: OutgoingJob) -> Self {
+        Self {
+            variant: Some(og.variant),
+        }
+    }
+}