Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

shorten apply-plan log #5214

Merged
merged 1 commit into from
Jul 12, 2024
Merged

shorten apply-plan log #5214

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 19 additions & 11 deletions quickwit/quickwit-common/src/pretty.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,28 +20,36 @@
use std::fmt;
use std::time::Duration;

pub struct PrettySample<'a, T>(&'a [T], usize);
pub struct PrettySample<I>(I, usize);

impl<'a, T> PrettySample<'a, T> {
pub fn new(slice: &'a [T], sample_size: usize) -> Self {
impl<I> PrettySample<I> {
pub fn new(slice: I, sample_size: usize) -> Self {
Self(slice, sample_size)
}
}

impl<T> fmt::Debug for PrettySample<'_, T>
where T: fmt::Debug
impl<I, T> fmt::Debug for PrettySample<I>
where
I: IntoIterator<Item = T> + Clone,
T: fmt::Debug,
{
fn fmt(&self, formatter: &mut fmt::Formatter<'_>) -> fmt::Result {
write!(formatter, "[")?;
for (i, item) in self.0.iter().enumerate() {
if i == self.1 {
write!(formatter, ", and {} more", self.0.len() - i)?;
break;
}
// in general we will get passed a reference (&[...], &HashMap...) or a Map<_> of them.
// So we either perform a Copy, or a cheap Clone of a simple struct
let mut iter = self.0.clone().into_iter().enumerate();
for (i, item) in &mut iter {
if i > 0 {
write!(formatter, ", ")?;
}
write!(formatter, "{item:?}")?;
if i == self.1 - 1 {
break;
}
}
let left = iter.count();
if left > 0 {
write!(formatter, ", and {left} more")?;
}
write!(formatter, "]")?;
Ok(())
Expand Down Expand Up @@ -83,7 +91,7 @@ mod tests {

#[test]
fn test_pretty_sample() {
let pretty_sample = PrettySample::<'_, usize>::new(&[], 2);
let pretty_sample = PrettySample::<&[usize]>::new(&[], 2);
assert_eq!(format!("{pretty_sample:?}"), "[]");

let pretty_sample = PrettySample::new(&[1], 2);
Expand Down
141 changes: 129 additions & 12 deletions quickwit/quickwit-control-plane/src/indexing_scheduler/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ use std::time::{Duration, Instant};
use fnv::{FnvHashMap, FnvHashSet};
use itertools::Itertools;
use once_cell::sync::OnceCell;
use quickwit_common::pretty::PrettySample;
use quickwit_proto::indexing::{
ApplyIndexingPlanRequest, CpuCapacity, IndexingService, IndexingTask, PIPELINE_FULL_CAPACITY,
PIPELINE_THROUGHTPUT,
Expand Down Expand Up @@ -448,38 +449,112 @@ fn get_shard_locality_metrics(
impl<'a> fmt::Debug for IndexingPlansDiff<'a> {
fn fmt(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
if self.has_same_nodes() && self.has_same_tasks() {
return write!(formatter, "EmptyIndexingPlanDiff");
return write!(formatter, "EmptyIndexingPlansDiff");
}
write!(formatter, "IndexingPlanDiff(")?;
write!(formatter, "IndexingPlansDiff(")?;
let mut separator = "";
if !self.missing_node_ids.is_empty() {
write!(formatter, "missing_node_ids={:?}, ", self.missing_node_ids)?;
write!(
formatter,
"missing_node_ids={:?}",
PrettySample::new(&self.missing_node_ids, 10)
)?;
separator = ", "
}
if !self.unplanned_node_ids.is_empty() {
write!(
formatter,
"{separator}unplanned_node_ids={:?}",
self.unplanned_node_ids
PrettySample::new(&self.unplanned_node_ids, 10)
)?;
separator = ", "
}
if !self.missing_tasks_by_node_id.is_empty() {
write!(
formatter,
"{separator}missing_tasks_by_node_id={:?}, ",
self.missing_tasks_by_node_id
)?;
write!(formatter, "{separator}missing_tasks_by_node_id=",)?;
format_indexing_task_map(formatter, &self.missing_tasks_by_node_id)?;
separator = ", "
}
if !self.unplanned_tasks_by_node_id.is_empty() {
write!(formatter, "{separator}unplanned_tasks_by_node_id=",)?;
format_indexing_task_map(formatter, &self.unplanned_tasks_by_node_id)?;
}
write!(formatter, ")")
}
}

fn format_indexing_task_map(
formatter: &mut std::fmt::Formatter,
indexing_tasks: &FnvHashMap<&str, Vec<&IndexingTask>>,
) -> std::fmt::Result {
// we show at most 5 nodes, and aggregate the results for the other.
// we show at most 10 indexes, but aggregate results after.
// we always aggregate shard ids
// we hide pipeline id and incarnation id, they are not very useful in most case, but take a
// lot of place
const MAX_NODE: usize = 5;
const MAX_INDEXES: usize = 10;
let mut index_displayed = 0;
write!(formatter, "{{")?;
let mut indexer_iter = indexing_tasks.iter().enumerate();
for (i, (index_name, tasks)) in &mut indexer_iter {
if i != 0 {
write!(formatter, ", ")?;
}
if index_displayed != MAX_INDEXES - 1 {
write!(formatter, "{index_name:?}: [")?;
let mut tasks_iter = tasks.iter().enumerate();
for (i, task) in &mut tasks_iter {
if i != 0 {
write!(formatter, ", ")?;
}
write!(
formatter,
r#"(index_id: "{}", source_id: "{}", shard_count: {})"#,
task.index_uid.as_ref().unwrap().index_id,
task.source_id,
task.shard_ids.len()
)?;
index_displayed += 1;
if index_displayed == MAX_INDEXES - 1 {
let (task_count, shard_count) = tasks_iter.fold((0, 0), |(t, s), (_, task)| {
(t + 1, s + task.shard_ids.len())
});
if task_count > 0 {
write!(
formatter,
" and {task_count} tasks and {shard_count} shards"
)?;
}
break;
}
}
write!(formatter, "]")?;
} else {
write!(
formatter,
"{separator}unplanned_tasks_by_node_id={:?}",
self.unplanned_tasks_by_node_id
"{index_name:?}: [with {} tasks and {} shards]",
tasks.len(),
tasks.iter().map(|task| task.shard_ids.len()).sum::<usize>()
)?;
}
write!(formatter, ")")
if i == MAX_NODE - 1 {
break;
}
}
let (indexer, tasks, shards) = indexer_iter.fold((0, 0, 0), |(i, t, s), (_, (_, task))| {
(
i + 1,
t + task.len(),
s + task.iter().map(|task| task.shard_ids.len()).sum::<usize>(),
)
});
if indexer > 0 {
write!(
formatter,
" and {indexer} more indexers, handling {tasks} tasks and {shards} shards}}"
)
} else {
write!(formatter, "}}")
}
}

Expand Down Expand Up @@ -884,6 +959,48 @@ mod tests {
assert_eq!(indexer_2_tasks.len(), 3);
}

#[test]
fn test_debug_indexing_task_map() {
let mut map = FnvHashMap::default();
let task1 = IndexingTask {
index_uid: Some(IndexUid::for_test("index1", 123)),
source_id: "my-source".to_string(),
pipeline_uid: Some(PipelineUid::random()),
shard_ids: vec!["shard1".into()],
};
let task2 = IndexingTask {
index_uid: Some(IndexUid::for_test("index2", 123)),
source_id: "my-source".to_string(),
pipeline_uid: Some(PipelineUid::random()),
shard_ids: vec!["shard2".into(), "shard3".into()],
};
let task3 = IndexingTask {
index_uid: Some(IndexUid::for_test("index3", 123)),
source_id: "my-source".to_string(),
pipeline_uid: Some(PipelineUid::random()),
shard_ids: vec!["shard6".into()],
};
// order made to map with the debug for lisibility
map.insert("indexer5", vec![&task2]);
map.insert("indexer4", vec![&task1]);
map.insert("indexer3", vec![&task1, &task3]);
map.insert("indexer2", vec![&task2, &task3, &task1, &task2]);
map.insert("indexer1", vec![&task1, &task2, &task3, &task1]);
map.insert("indexer6", vec![&task1, &task2, &task3]);
let plan = IndexingPlansDiff {
missing_node_ids: FnvHashSet::default(),
unplanned_node_ids: FnvHashSet::default(),
missing_tasks_by_node_id: map,
unplanned_tasks_by_node_id: FnvHashMap::default(),
};

let debug = format!("{plan:?}");
assert_eq!(
debug,
r#"IndexingPlansDiff(missing_tasks_by_node_id={"indexer5": [(index_id: "index2", source_id: "my-source", shard_count: 2)], "indexer4": [(index_id: "index1", source_id: "my-source", shard_count: 1)], "indexer3": [(index_id: "index1", source_id: "my-source", shard_count: 1), (index_id: "index3", source_id: "my-source", shard_count: 1)], "indexer2": [(index_id: "index2", source_id: "my-source", shard_count: 2), (index_id: "index3", source_id: "my-source", shard_count: 1), (index_id: "index1", source_id: "my-source", shard_count: 1), (index_id: "index2", source_id: "my-source", shard_count: 2)], "indexer1": [(index_id: "index1", source_id: "my-source", shard_count: 1) and 3 tasks and 4 shards] and 1 more indexers, handling 3 tasks and 4 shards})"#
);
}

proptest! {
#[test]
fn test_building_indexing_tasks_and_physical_plan(num_indexers in 1usize..50usize, index_id_sources in proptest::collection::vec(gen_kafka_source(), 1..20)) {
Expand Down
Loading