page_service: rewrite batching to work without a timeout (#9851)

# Problem

The timeout-based batching adds latency to unbatchable workloads.

We can choose a short batching timeout (e.g. 10us) but that requires
high-resolution timers, which tokio doesn't have.
I thoroughly explored options to use OS timers (see
[this](#9822) abandoned PR).
In short, it's not an attractive option because any timer implementation
adds non-trivial overheads.

# Solution

The insight is that, in the steady state of a batchable workload, the
time we spend in `get_vectored` will be hundreds of microseconds anyway.

If we prepare the next batch concurrently to `get_vectored`, we will
have a sizeable batch ready once `get_vectored` of the current batch is
done and do not need an explicit timeout.

This can be reasonably described as **pipelining of the protocol
handler**.

# Implementation

We model the sub-protocol handler for pagestream requests
(`handle_pagrequests`) as two futures that form a pipeline:

2. Batching: read requests from the connection and fill the current
batch
3. Execution: `take` the current batch, execute it using `get_vectored`,
and send the response.

The Reading and Batching stage are connected through a new type of
channel called `spsc_fold`.

See the long comment in the `handle_pagerequests_pipelined` for details.

# Changes

- Refactor `handle_pagerequests`
    - separate functions for
- reading one protocol message; produces a `BatchedFeMessage` with just
one page request in it
- batching; tried to merge an incoming `BatchedFeMessage` into an
existing `BatchedFeMessage`; returns `None` on success and returns back
the incoming message in case merging isn't possible
        - execution of a batched message
- unify the timeline handle acquisition & request span construction; it
now happen in the function that reads the protocol message
- Implement serial and pipelined model
    - serial: what we had before any of the batching changes
      - read one protocol message
      - execute protocol messages
    - pipelined: the design described above
- optionality for execution of the pipeline: either via concurrent
futures vs tokio tasks
- Pageserver config
  - remove batching timeout field
  - add ability to configure pipelining mode
- add ability to limit max batch size for pipelined configurations
(required for the rollout, cf
neondatabase/cloud#20620 )
  - ability to configure execution mode
- Tests
  - remove `batch_timeout` parametrization
  - rename `test_getpage_merge_smoke` to `test_throughput`
- add parametrization to test different max batch sizes and execution
moes
  - rename `test_timer_precision` to `test_latency`
  - rename the test case file to `test_page_service_batching.py`
  - better descriptions of what the tests actually do

## On the holding The `TimelineHandle` in the pending batch

While batching, we hold the `TimelineHandle` in the pending batch.
Therefore, the timeline will not finish shutting down while we're
batching.

This is not a problem in practice because the concurrently ongoing
`get_vectored` call will fail quickly with an error indicating that the
timeline is shutting down.
This results in the Execution stage returning a `QueryError::Shutdown`,
which causes the pipeline / entire page service connection to shut down.
This drops all references to the
`Arc<Mutex<Option<Box<BatchedFeMessage>>>>` object, thereby dropping the
contained `TimelineHandle`s.

- => fixes #9850

# Performance

Local run of the benchmarks, results in [this empty
commit](1cf5b14)
in the PR branch.

Key take-aways:
* `concurrent-futures` and `tasks` deliver identical `batching_factor`
* tail latency impact unknown, cf
#9837
* `concurrent-futures` has higher throughput than `tasks` in all
workloads (=lower `time` metric)
* In unbatchable workloads, `concurrent-futures` has 5% higher
`CPU-per-throughput` than that of `tasks`, and 15% higher than that of
`serial`.
* In batchable-32 workload, `concurrent-futures` has 8% lower
`CPU-per-throughput` than that of `tasks` (comparison to tput of
`serial` is irrelevant)
* in unbatchable workloads, mean and tail latencies of
`concurrent-futures` is practically identical to `serial`, whereas
`tasks` adds 20-30us of overhead

Overall, `concurrent-futures` seems like a slightly more attractive
choice.

# Rollout

This change is disabled-by-default.

Rollout plan:
- neondatabase/cloud#20620

# Refs

- epic: #9376
- this sub-task: #9377
- the abandoned attempt to improve batching timeout resolution:
#9820
- closes #9850
- fixes #9835

Cargo.toml

@@ -83,6 +83,7 @@ comfy-table = "7.1"
 const_format = "0.2"
 crc32c = "0.6"
 dashmap = { version = "5.5.0", features = ["raw-api"] }
+diatomic-waker = { version = "0.2.3" }
 either = "1.8"
 enum-map = "2.4.2"
 enumset = "1.0.12"

libs/pageserver_api/src/config.rs

@@ -118,9 +118,8 @@ pub struct ConfigToml {
     pub virtual_file_io_mode: Option<crate::models::virtual_file::IoMode>,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub no_sync: Option<bool>,
-    #[serde(with = "humantime_serde")]
-    pub server_side_batch_timeout: Option<Duration>,
     pub wal_receiver_protocol: PostgresClientProtocol,
+    pub page_service_pipelining: PageServicePipeliningConfig,
 }
 
 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
@@ -137,6 +136,28 @@ pub struct DiskUsageEvictionTaskConfig {
     pub eviction_order: EvictionOrder,
 }
 
+#[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
+#[serde(tag = "mode", rename_all = "kebab-case")]
+#[serde(deny_unknown_fields)]
+pub enum PageServicePipeliningConfig {
+    Serial,
+    Pipelined(PageServicePipeliningConfigPipelined),
+}
+#[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
+#[serde(deny_unknown_fields)]
+pub struct PageServicePipeliningConfigPipelined {
+    /// Causes runtime errors if larger than max get_vectored batch size.
+    pub max_batch_size: NonZeroUsize,
+    pub execution: PageServiceProtocolPipelinedExecutionStrategy,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
+#[serde(rename_all = "kebab-case")]
+pub enum PageServiceProtocolPipelinedExecutionStrategy {
+    ConcurrentFutures,
+    Tasks,
+}
+
 pub mod statvfs {
     pub mod mock {
         #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
@@ -332,8 +353,6 @@ pub mod defaults {
 
     pub const DEFAULT_IO_BUFFER_ALIGNMENT: usize = 512;
 
-    pub const DEFAULT_SERVER_SIDE_BATCH_TIMEOUT: Option<&str> = None;
-
     pub const DEFAULT_WAL_RECEIVER_PROTOCOL: utils::postgres_client::PostgresClientProtocol =
         utils::postgres_client::PostgresClientProtocol::Vanilla;
 }
@@ -420,11 +439,10 @@ impl Default for ConfigToml {
             ephemeral_bytes_per_memory_kb: (DEFAULT_EPHEMERAL_BYTES_PER_MEMORY_KB),
             l0_flush: None,
             virtual_file_io_mode: None,
-            server_side_batch_timeout: DEFAULT_SERVER_SIDE_BATCH_TIMEOUT
-                .map(|duration| humantime::parse_duration(duration).unwrap()),
             tenant_config: TenantConfigToml::default(),
             no_sync: None,
             wal_receiver_protocol: DEFAULT_WAL_RECEIVER_PROTOCOL,
+            page_service_pipelining: PageServicePipeliningConfig::Serial,
         }
     }
 }

libs/utils/Cargo.toml

@@ -19,6 +19,7 @@ bincode.workspace = true
 bytes.workspace = true
 camino.workspace = true
 chrono.workspace = true
+diatomic-waker.workspace = true
 git-version.workspace = true
 hex = { workspace = true, features = ["serde"] }
 humantime.workspace = true
@@ -45,6 +46,7 @@ tracing.workspace = true
 tracing-error.workspace = true
 tracing-subscriber = { workspace = true, features = ["json", "registry"] }
 rand.workspace = true
+scopeguard.workspace = true
 strum.workspace = true
 strum_macros.workspace = true
 url.workspace = true

libs/utils/src/sync.rs

@@ -1,3 +1,5 @@
 pub mod heavier_once_cell;
 
 pub mod gate;
+
+pub mod spsc_fold;