Request based PoV distribution (#2640)

* Indentation fix.

* Prepare request-response for PoV fetching.

* Drop old PoV distribution.

* WIP: Fetch PoV directly from backing.

* Backing compiles.

* Runtime access and connection management for PoV distribution.

* Get rid of seemingly dead code.

* Implement PoV fetching.

Backing does not yet use it.

* Don't send `ConnectToValidators` for empty list.

* Even better - no need to check over and over again.

* PoV fetching implemented.

+ Typechecks
+ Should work

Missing:

- Guide
- Tests
- Do fallback fetching in case fetching from seconding validator fails.

* Check PoV hash upon reception.

* Implement retry of PoV fetching in backing.

* Avoid pointless validation spawning.

* Add jaeger span to pov requesting.

* Add back tracing.

* Review remarks.

* Whitespace.

* Whitespace again.

* Cleanup + fix tests.

* Log to log target in overseer.

* Fix more tests.

* Don't fail if group cannot be found.

* Simple test for PoV fetcher.

* Handle missing group membership better.

* Add test for retry functionality.

* Fix flaky test.

* Spaces again.

* Guide updates.

* Spaces.
This commit is contained in:
Robert Klotzner
2021-03-28 17:11:38 +02:00
committed by GitHub
parent 27b6d83974
commit c6f07d8f31
35 changed files with 1382 additions and 3184 deletions
@@ -39,7 +39,7 @@ use polkadot_subsystem::{
};
use super::{error::recv_runtime, session_cache::SessionCache, LOG_TARGET, Metrics};
use crate::error::NonFatalError;
use crate::error::Error;
/// A task fetching a particular chunk.
mod fetch_task;
@@ -96,7 +96,7 @@ impl Requester {
&mut self,
ctx: &mut Context,
update: ActiveLeavesUpdate,
) -> super::Result<Option<NonFatalError>>
) -> super::Result<()>
where
Context: SubsystemContext,
{
@@ -111,9 +111,9 @@ impl Requester {
} = update;
// Order important! We need to handle activated, prior to deactivated, otherwise we might
// cancel still needed jobs.
let err = self.start_requesting_chunks(ctx, activated.into_iter()).await?;
self.start_requesting_chunks(ctx, activated.into_iter()).await?;
self.stop_requesting_chunks(deactivated.into_iter());
Ok(err)
Ok(())
}
/// Start requesting chunks for newly imported heads.
@@ -121,25 +121,20 @@ impl Requester {
&mut self,
ctx: &mut Context,
new_heads: impl Iterator<Item = ActivatedLeaf>,
) -> super::Result<Option<NonFatalError>>
) -> super::Result<()>
where
Context: SubsystemContext,
{
for ActivatedLeaf { hash: leaf, .. } in new_heads {
let cores = match query_occupied_cores(ctx, leaf).await {
Err(err) => return Ok(Some(err)),
Ok(cores) => cores,
};
let cores = query_occupied_cores(ctx, leaf).await?;
tracing::trace!(
target: LOG_TARGET,
occupied_cores = ?cores,
"Query occupied core"
);
if let Some(err) = self.add_cores(ctx, leaf, cores).await? {
return Ok(Some(err));
}
self.add_cores(ctx, leaf, cores).await?;
}
Ok(None)
Ok(())
}
/// Stop requesting chunks for obsolete heads.
@@ -164,7 +159,7 @@ impl Requester {
ctx: &mut Context,
leaf: Hash,
cores: impl IntoIterator<Item = OccupiedCore>,
) -> super::Result<Option<NonFatalError>>
) -> super::Result<()>
where
Context: SubsystemContext,
{
@@ -179,7 +174,7 @@ impl Requester {
let tx = self.tx.clone();
let metrics = self.metrics.clone();
let task_cfg = match self
let task_cfg = self
.session_cache
.with_session_info(
ctx,
@@ -189,11 +184,7 @@ impl Requester {
leaf,
|info| FetchTaskConfig::new(leaf, &core, tx, metrics, info),
)
.await
{
Err(err) => return Ok(Some(err)),
Ok(task_cfg) => task_cfg,
};
.await?;
if let Some(task_cfg) = task_cfg {
e.insert(FetchTask::start(task_cfg, ctx).await?);
@@ -202,7 +193,7 @@ impl Requester {
}
}
}
Ok(None)
Ok(())
}
}
@@ -237,7 +228,7 @@ impl Stream for Requester {
async fn query_occupied_cores<Context>(
ctx: &mut Context,
relay_parent: Hash,
) -> Result<Vec<OccupiedCore>, NonFatalError>
) -> Result<Vec<OccupiedCore>, Error>
where
Context: SubsystemContext,
{