Merge remote-tracking branch 'origin/main' into problame/hung-shutdown/demo-hypothesis

libs/pageserver_api/src/controller_api.rs

@@ -87,7 +87,7 @@ impl Display for AvailabilityZone {
 #[derive(Serialize, Deserialize)]
 pub struct ShardsPreferredAzsRequest {
     #[serde(flatten)]
-    pub preferred_az_ids: HashMap<TenantShardId, AvailabilityZone>,
+    pub preferred_az_ids: HashMap<TenantShardId, Option<AvailabilityZone>>,
 }
 
 #[derive(Serialize, Deserialize)]
@@ -144,6 +144,8 @@ pub struct NodeDescribeResponse {
     pub availability: NodeAvailabilityWrapper,
     pub scheduling: NodeSchedulingPolicy,
 
+    pub availability_zone_id: String,
+
     pub listen_http_addr: String,
     pub listen_http_port: u16,
 
@@ -179,7 +181,6 @@ pub struct TenantDescribeResponseShard {
 /// specifies some constraints, e.g. asking it to get off particular node(s)
 #[derive(Serialize, Deserialize, Debug)]
 pub struct TenantShardMigrateRequest {
-    pub tenant_shard_id: TenantShardId,
     pub node_id: NodeId,
 }
 
@@ -368,6 +369,16 @@ pub enum PlacementPolicy {
     Detached,
 }
 
+impl PlacementPolicy {
+    pub fn want_secondaries(&self) -> usize {
+        match self {
+            PlacementPolicy::Attached(secondary_count) => *secondary_count,
+            PlacementPolicy::Secondary => 1,
+            PlacementPolicy::Detached => 0,
+        }
+    }
+}
+
 #[derive(Serialize, Deserialize, Debug)]
 pub struct TenantShardMigrateResponse {}
 

libs/pageserver_api/src/key.rs

@@ -706,7 +706,7 @@ pub fn repl_origin_key_range() -> Range<Key> {
 /// Non inherited range for vectored get.
 pub const NON_INHERITED_RANGE: Range<Key> = AUX_FILES_KEY..AUX_FILES_KEY.next();
 /// Sparse keyspace range for vectored get. Missing key error will be ignored for this range.
-pub const NON_INHERITED_SPARSE_RANGE: Range<Key> = Key::metadata_key_range();
+pub const SPARSE_RANGE: Range<Key> = Key::metadata_key_range();
 
 impl Key {
     // AUX_FILES currently stores only data for logical replication (slots etc), and
@@ -714,7 +714,42 @@ impl Key {
     // switch (and generally it likely should be optional), so ignore these.
     #[inline(always)]
     pub fn is_inherited_key(self) -> bool {
-        !NON_INHERITED_RANGE.contains(&self) && !NON_INHERITED_SPARSE_RANGE.contains(&self)
+        if self.is_sparse() {
+            self.is_inherited_sparse_key()
+        } else {
+            !NON_INHERITED_RANGE.contains(&self)
+        }
+    }
+
+    #[inline(always)]
+    pub fn is_sparse(self) -> bool {
+        self.field1 >= METADATA_KEY_BEGIN_PREFIX && self.field1 < METADATA_KEY_END_PREFIX
+    }
+
+    /// Check if the key belongs to the inherited keyspace.
+    fn is_inherited_sparse_key(self) -> bool {
+        debug_assert!(self.is_sparse());
+        self.field1 == RELATION_SIZE_PREFIX
+    }
+
+    pub fn sparse_non_inherited_keyspace() -> Range<Key> {
+        // The two keys are adjacent; if we will have non-adjancent keys in the future, we should return a keyspace
+        debug_assert_eq!(AUX_KEY_PREFIX + 1, REPL_ORIGIN_KEY_PREFIX);
+        Key {
+            field1: AUX_KEY_PREFIX,
+            field2: 0,
+            field3: 0,
+            field4: 0,
+            field5: 0,
+            field6: 0,
+        }..Key {
+            field1: REPL_ORIGIN_KEY_PREFIX + 1,
+            field2: 0,
+            field3: 0,
+            field4: 0,
+            field5: 0,
+            field6: 0,
+        }
     }
 
     #[inline(always)]

libs/pageserver_api/src/models.rs

@@ -272,6 +272,8 @@ pub struct CompactInfoResponse {
     pub compact_key_range: Option<CompactKeyRange>,
     pub compact_lsn_range: Option<CompactLsnRange>,
     pub sub_compaction: bool,
+    pub running: bool,
+    pub job_id: usize,
 }
 
 #[derive(Serialize, Deserialize, Clone)]

libs/pq_proto/src/lib.rs

@@ -44,7 +44,7 @@ pub struct ProtocolVersion(u32);
 
 impl ProtocolVersion {
     pub const fn new(major: u16, minor: u16) -> Self {
-        Self((major as u32) << 16 | minor as u32)
+        Self(((major as u32) << 16) | minor as u32)
     }
     pub const fn minor(self) -> u16 {
         self.0 as u16

libs/remote_storage/src/config.rs

@@ -43,6 +43,17 @@ impl RemoteStorageKind {
     }
 }
 
+impl RemoteStorageConfig {
+    /// Helper to fetch the configured concurrency limit.
+    pub fn concurrency_limit(&self) -> Option<usize> {
+        match &self.storage {
+            RemoteStorageKind::LocalFs { .. } => None,
+            RemoteStorageKind::AwsS3(c) => Some(c.concurrency_limit.into()),
+            RemoteStorageKind::AzureContainer(c) => Some(c.concurrency_limit.into()),
+        }
+    }
+}
+
 fn default_timeout() -> Duration {
     RemoteStorageConfig::DEFAULT_TIMEOUT
 }

libs/utils/src/generation.rs

@@ -112,9 +112,9 @@ impl Serialize for Generation {
             // We should never be asked to serialize a None. Structures
             // that include an optional generation should convert None to an
             // Option<Generation>::None
-            Err(serde::ser::Error::custom(
-                "Tried to serialize invalid generation ({self})",
-            ))
+            Err(serde::ser::Error::custom(format!(
+                "Tried to serialize invalid generation ({self:?})"
+            )))
         }
     }
 }

libs/utils/src/http/endpoint.rs

@@ -15,7 +15,7 @@ use once_cell::sync::Lazy;
 use regex::Regex;
 use routerify::ext::RequestExt;
 use routerify::{Middleware, RequestInfo, Router, RouterBuilder};
-use tokio::sync::{mpsc, Mutex};
+use tokio::sync::{mpsc, Mutex, Notify};
 use tokio_stream::wrappers::ReceiverStream;
 use tokio_util::io::ReaderStream;
 use tracing::{debug, info, info_span, warn, Instrument};
@@ -350,33 +350,53 @@ pub async fn profile_cpu_handler(req: Request<Body>) -> Result<Response<Body>, A
     };
     let seconds = match parse_query_param(&req, "seconds")? {
         None => 5,
-        Some(seconds @ 1..=30) => seconds,
-        Some(_) => return Err(ApiError::BadRequest(anyhow!("duration must be 1-30 secs"))),
+        Some(seconds @ 1..=60) => seconds,
+        Some(_) => return Err(ApiError::BadRequest(anyhow!("duration must be 1-60 secs"))),
     };
     let frequency_hz = match parse_query_param(&req, "frequency")? {
         None => 99,
         Some(1001..) => return Err(ApiError::BadRequest(anyhow!("frequency must be <=1000 Hz"))),
         Some(frequency) => frequency,
     };
-
-    // Only allow one profiler at a time.
-    static PROFILE_LOCK: Lazy<Mutex<()>> = Lazy::new(|| Mutex::new(()));
-    let _lock = PROFILE_LOCK
-        .try_lock()
-        .map_err(|_| ApiError::Conflict("profiler already running".into()))?;
+    let force: bool = parse_query_param(&req, "force")?.unwrap_or_default();
 
     // Take the profile.
-    let report = tokio::task::spawn_blocking(move || {
+    static PROFILE_LOCK: Lazy<Mutex<()>> = Lazy::new(|| Mutex::new(()));
+    static PROFILE_CANCEL: Lazy<Notify> = Lazy::new(Notify::new);
+
+    let report = {
+        // Only allow one profiler at a time. If force is true, cancel a running profile (e.g. a
+        // Grafana continuous profile). We use a try_lock() loop when cancelling instead of waiting
+        // for a lock(), to avoid races where the notify isn't currently awaited.
+        let _lock = loop {
+            match PROFILE_LOCK.try_lock() {
+                Ok(lock) => break lock,
+                Err(_) if force => PROFILE_CANCEL.notify_waiters(),
+                Err(_) => {
+                    return Err(ApiError::Conflict(
+                        "profiler already running (use ?force=true to cancel it)".into(),
+                    ))
+                }
+            }
+            tokio::time::sleep(Duration::from_millis(1)).await; // don't busy-wait
+        };
+
         let guard = ProfilerGuardBuilder::default()
             .frequency(frequency_hz)
             .blocklist(&["libc", "libgcc", "pthread", "vdso"])
-            .build()?;
-        std::thread::sleep(Duration::from_secs(seconds));
-        guard.report().build()
-    })
-    .await
-    .map_err(|join_err| ApiError::InternalServerError(join_err.into()))?
-    .map_err(|pprof_err| ApiError::InternalServerError(pprof_err.into()))?;
+            .build()
+            .map_err(|err| ApiError::InternalServerError(err.into()))?;
+
+        tokio::select! {
+            _ = tokio::time::sleep(Duration::from_secs(seconds)) => {},
+            _ = PROFILE_CANCEL.notified() => {},
+        };
+
+        guard
+            .report()
+            .build()
+            .map_err(|err| ApiError::InternalServerError(err.into()))?
+    };
 
     // Return the report in the requested format.
     match format {

libs/utils/src/lsn.rs

@@ -260,7 +260,7 @@ impl FromStr for Lsn {
         {
             let left_num = u32::from_str_radix(left, 16).map_err(|_| LsnParseError)?;
             let right_num = u32::from_str_radix(right, 16).map_err(|_| LsnParseError)?;
-            Ok(Lsn((left_num as u64) << 32 | right_num as u64))
+            Ok(Lsn(((left_num as u64) << 32) | right_num as u64))
         } else {
             Err(LsnParseError)
         }