mz_compute/render/
sinks.rs

1// Copyright Materialize, Inc. and contributors. All rights reserved.
2//
3// Use of this software is governed by the Business Source License
4// included in the LICENSE file.
5//
6// As of the Change Date specified in that file, in accordance with
7// the Business Source License, use of this software will be governed
8// by the Apache License, Version 2.0.
9
10//! Logic related to the creation of dataflow sinks.
11
12use std::any::Any;
13use std::collections::{BTreeMap, BTreeSet};
14use std::rc::Rc;
15
16use differential_dataflow::VecCollection;
17use mz_compute_types::plan::scalar::LirScalarExpr;
18use mz_compute_types::sinks::{ComputeSinkConnection, ComputeSinkDesc};
19use mz_expr::{EvalError, MapFilterProject, permutation_for_arrangement};
20use mz_ore::soft_assert_or_log;
21use mz_ore::str::StrExt;
22use mz_ore::vec::PartialOrdVecExt;
23use mz_repr::{Diff, GlobalId, Row};
24use mz_storage_types::controller::CollectionMetadata;
25use mz_timely_util::operator::CollectionExt;
26use mz_timely_util::probe::Handle;
27use timely::container::CapacityContainerBuilder;
28use timely::dataflow::Scope;
29use timely::progress::Antichain;
30
31use crate::compute_state::SinkToken;
32use crate::logging::compute::LogDataflowErrors;
33use crate::render::context::Context;
34use crate::render::errors::DataflowErrorSer;
35use crate::render::{RenderTimestamp, StartSignal};
36
37impl<'g, T: RenderTimestamp> Context<'g, T> {
38    /// Export the sink described by `sink` from the rendering context.
39    pub(crate) fn export_sink(
40        &self,
41        compute_state: &mut crate::compute_state::ComputeState,
42        tokens: &BTreeMap<GlobalId, Rc<dyn Any>>,
43        dependency_ids: BTreeSet<GlobalId>,
44        sink_id: GlobalId,
45        sink: &ComputeSinkDesc<CollectionMetadata>,
46        start_signal: StartSignal,
47        output_probe: &Handle<mz_repr::Timestamp>,
48        outer_scope: Scope<'g, mz_repr::Timestamp>,
49    ) {
50        soft_assert_or_log!(
51            sink.non_null_assertions.is_strictly_sorted(),
52            "non-null assertions not sorted"
53        );
54
55        // put together tokens that belong to the export
56        let mut needed_tokens = Vec::new();
57        for dep_id in dependency_ids {
58            if let Some(token) = tokens.get(&dep_id) {
59                needed_tokens.push(Rc::clone(token))
60            }
61        }
62
63        // TODO[btv] - We should determine the key and permutation to use during planning,
64        // rather than at runtime.
65        //
66        // This is basically an inlined version of the old `as_collection`.
67        let bundle = self
68            .lookup_id(mz_expr::Id::Global(sink.from))
69            .expect("Sink source collection not loaded");
70        let (ok_collection, mut err_collection) = if let Some(collection) = &bundle.collection {
71            collection.clone()
72        } else {
73            let (key, _arrangement) = bundle
74                .arranged
75                .iter()
76                .next()
77                .expect("Invariant violated: at least one collection must be present.");
78            let unthinned_arity = sink.from_desc.arity();
79            let (permutation, thinning) = permutation_for_arrangement(key, unthinned_arity);
80            let mut mfp = MapFilterProject::<LirScalarExpr>::new(unthinned_arity);
81            mfp.permute_fn(|c| permutation[c], thinning.len() + key.len());
82            let mfp_plan = mfp.into_plan().expect("MFP planning failed");
83            bundle.as_collection_core(
84                mfp_plan,
85                Some((key.clone(), None)),
86                self.until.clone(),
87                &self.config_set,
88            )
89        };
90
91        // Attach logging of dataflow errors.
92        if let Some(logger) = compute_state.compute_logger.clone() {
93            err_collection = err_collection.log_dataflow_errors(logger, sink_id);
94        }
95
96        let mut ok_collection = ok_collection.leave(outer_scope);
97        let mut err_collection = err_collection.leave(outer_scope);
98
99        // Ensure that the frontier does not advance past the expiration time, if set. Otherwise,
100        // we might write down incorrect data.
101        if let Some(&expiration) = self.dataflow_expiration.as_option() {
102            ok_collection = ok_collection
103                .expire_collection_at(&format!("{}_export_sink_oks", self.debug_name), expiration);
104            err_collection = err_collection
105                .expire_collection_at(&format!("{}_export_sink_errs", self.debug_name), expiration);
106        }
107
108        let non_null_assertions = sink.non_null_assertions.clone();
109        let from_desc = sink.from_desc.clone();
110        if !non_null_assertions.is_empty() {
111            let name = format!("NullAssertions({sink_id:?})");
112            type CB<C> = CapacityContainerBuilder<C>;
113            let (oks, null_errs) =
114                ok_collection.map_fallible::<CB<_>, CB<_>, _, _, _>(&name, move |row| {
115                    let mut idx = 0;
116                    let mut iter = row.iter();
117                    for &i in &non_null_assertions {
118                        let skip = i - idx;
119                        let datum = iter.nth(skip).unwrap();
120                        idx += skip + 1;
121                        if datum.is_null() {
122                            return Err(DataflowErrorSer::from(EvalError::MustNotBeNull(
123                                format!("column {}", from_desc.get_name(i).quoted()).into(),
124                            )));
125                        }
126                    }
127                    Ok(row)
128                });
129            ok_collection = oks;
130            err_collection = err_collection.concat(null_errs);
131        }
132
133        let region_name = match sink.connection {
134            ComputeSinkConnection::Subscribe(_) => format!("SubscribeSink({:?})", sink_id),
135            ComputeSinkConnection::MaterializedView(_) => {
136                format!("MaterializedViewSink({:?})", sink_id)
137            }
138            ComputeSinkConnection::CopyToS3Oneshot(_) => {
139                format!("CopyToS3OneshotSink({:?})", sink_id)
140            }
141        };
142        outer_scope.clone().region_named(&region_name, |inner| {
143            let sink_render = get_sink_render_for(&sink.connection);
144
145            let sink_token = sink_render.render_sink(
146                compute_state,
147                sink,
148                sink_id,
149                self.as_of_frontier.clone(),
150                start_signal,
151                ok_collection.enter_region(inner),
152                err_collection.enter_region(inner),
153                output_probe,
154            );
155
156            if let Some(sink_token) = sink_token {
157                needed_tokens.push(sink_token);
158            }
159
160            let collection = compute_state.expect_collection_mut(sink_id);
161            collection.sink_token = Some(SinkToken::new(Box::new(needed_tokens)));
162        });
163    }
164}
165
166/// A type that can be rendered as a dataflow sink.
167pub(crate) trait SinkRender<'scope> {
168    fn render_sink(
169        &self,
170        compute_state: &mut crate::compute_state::ComputeState,
171        sink: &ComputeSinkDesc<CollectionMetadata>,
172        sink_id: GlobalId,
173        as_of: Antichain<mz_repr::Timestamp>,
174        start_signal: StartSignal,
175        sinked_collection: VecCollection<'scope, mz_repr::Timestamp, Row, Diff>,
176        err_collection: VecCollection<'scope, mz_repr::Timestamp, DataflowErrorSer, Diff>,
177        output_probe: &Handle<mz_repr::Timestamp>,
178    ) -> Option<Rc<dyn Any>>;
179}
180
181fn get_sink_render_for<'scope>(
182    connection: &ComputeSinkConnection<CollectionMetadata>,
183) -> Box<dyn SinkRender<'scope>> {
184    match connection {
185        ComputeSinkConnection::Subscribe(connection) => Box::new(connection.clone()),
186        ComputeSinkConnection::MaterializedView(connection) => Box::new(connection.clone()),
187        ComputeSinkConnection::CopyToS3Oneshot(connection) => Box::new(connection.clone()),
188    }
189}
mz_compute/render/sinks.rs

mz_compute/render/
sinks.rs
