[RSS] Move switch_port table population to a Nexus bg task

clients/nexus-lockstep-client/src/lib.rs

@@ -50,7 +50,6 @@ progenitor::generate_api!(
         DnsConfigParams = nexus_types::internal_api::params::DnsConfigParams,
         DnsConfigZone = nexus_types::internal_api::params::DnsConfigZone,
         DnsRecord = nexus_types::internal_api::params::DnsRecord,
-        ExternalPortDiscovery = nexus_types::internal_api::params::ExternalPortDiscovery,
         Generation = omicron_common::api::external::Generation,
         ImportExportPolicy = sled_agent_types::early_networking::ImportExportPolicy,
         MacAddr = omicron_common::api::external::MacAddr,

dev-tools/omdb/src/bin/omdb/nexus.rs

@@ -81,6 +81,8 @@ use nexus_types::internal_api::background::SupportBundleCleanupReport;
 use nexus_types::internal_api::background::SupportBundleCollectionReport;
 use nexus_types::internal_api::background::SupportBundleCollectionStepStatus;
 use nexus_types::internal_api::background::SupportBundleEreportStatus;
+use nexus_types::internal_api::background::SwitchPortPopulatorStatus;
+use nexus_types::internal_api::background::SwitchPortPopulatorStatusKind;
 use nexus_types::internal_api::background::TrustQuorumManagerStatus;
 use nexus_types::internal_api::background::TufArtifactReplicationCounters;
 use nexus_types::internal_api::background::TufArtifactReplicationRequest;
@@ -1369,6 +1371,9 @@ fn print_task_details(bgtask: &BackgroundTask, details: &serde_json::Value) {
         "trust_quorum_manager" => {
             print_task_trust_quorum_manager(details);
         }
+        "populate_switch_ports" => {
+            print_task_populate_switch_ports(details);
+        }
         _ => {
             println!(
                 "warning: unknown background task: {:?} \
@@ -3940,6 +3945,40 @@ fn print_task_trust_quorum_manager(details: &serde_json::Value) {
     }
 }
 
+fn print_task_populate_switch_ports(details: &serde_json::Value) {
+    fn print_one(
+        name: &str,
+        result: Result<SwitchPortPopulatorStatusKind, String>,
+    ) {
+        match result {
+            Ok(SwitchPortPopulatorStatusKind::Populated { num_ports }) => {
+                println!("{name}: populated {num_ports} ports");
+            }
+            Ok(SwitchPortPopulatorStatusKind::PreviouslyPopulated) => {
+                println!("{name} skipped: previously populated ports");
+            }
+            Err(err) => println!("{name} failed: {err}"),
+        }
+    }
+
+    let status = match serde_json::from_value::<SwitchPortPopulatorStatus>(
+        details.clone(),
+    ) {
+        Ok(status) => status,
+        Err(error) => {
+            eprintln!(
+                "warning: failed to interpret task details: {:?}: {:#?}",
+                error, details
+            );
+            return;
+        }
+    };
+
+    let SwitchPortPopulatorStatus { switch0, switch1 } = status;
+    print_one("switch0", switch0);
+    print_one("switch1", switch1);
+}
+
 const ERRICON: &str = "/!\\";
 
 fn warn_if_nonzero(n: usize) -> &'static str {

dev-tools/omdb/tests/env.out

@@ -176,6 +176,11 @@ task: "physical_disk_adoption"
     ensure new physical disks are automatically marked in-service
 
 
+task: "populate_switch_ports"
+    one-time population of the `switch_port` table containing all QSFP ports
+    managed by dendrite
+
+
 task: "probe_distributor"
     distributes networking probe zones to sleds
 
@@ -439,6 +444,11 @@ task: "physical_disk_adoption"
     ensure new physical disks are automatically marked in-service
 
 
+task: "populate_switch_ports"
+    one-time population of the `switch_port` table containing all QSFP ports
+    managed by dendrite
+
+
 task: "probe_distributor"
     distributes networking probe zones to sleds
 
@@ -689,6 +699,11 @@ task: "physical_disk_adoption"
     ensure new physical disks are automatically marked in-service
 
 
+task: "populate_switch_ports"
+    one-time population of the `switch_port` table containing all QSFP ports
+    managed by dendrite
+
+
 task: "probe_distributor"
     distributes networking probe zones to sleds
 

dev-tools/omdb/tests/successes.out

@@ -411,6 +411,11 @@ task: "physical_disk_adoption"
     ensure new physical disks are automatically marked in-service
 
 
+task: "populate_switch_ports"
+    one-time population of the `switch_port` table containing all QSFP ports
+    managed by dendrite
+
+
 task: "probe_distributor"
     distributes networking probe zones to sleds
 
@@ -854,6 +859,13 @@ task: "physical_disk_adoption"
     started at <REDACTED_TIMESTAMP> (<REDACTED DURATION>s ago) and ran for <REDACTED DURATION>ms
     last completion reported error: task disabled
 
+task: "populate_switch_ports"
+  configured period: every <REDACTED_DURATION>s
+  last completed activation: <REDACTED ITERATIONS>, triggered by <TRIGGERED_BY_REDACTED>
+    started at <REDACTED_TIMESTAMP> (<REDACTED DURATION>s ago) and ran for <REDACTED DURATION>ms
+switch0 failed: failed to look up dendrite clients: proto error: no records found for Query { name: Name("_dendrite._tcp.control-plane.oxide.internal."), query_type: SRV, query_class: IN }
+switch1 failed: failed to look up dendrite clients: proto error: no records found for Query { name: Name("_dendrite._tcp.control-plane.oxide.internal."), query_type: SRV, query_class: IN }
+
 task: "probe_distributor"
   configured period: every <REDACTED_DURATION>m
   last completed activation: <REDACTED ITERATIONS>, triggered by <TRIGGERED_BY_REDACTED>
@@ -1535,6 +1547,13 @@ task: "physical_disk_adoption"
     started at <REDACTED_TIMESTAMP> (<REDACTED DURATION>s ago) and ran for <REDACTED DURATION>ms
     last completion reported error: task disabled
 
+task: "populate_switch_ports"
+  configured period: every <REDACTED_DURATION>s
+  last completed activation: <REDACTED ITERATIONS>, triggered by <TRIGGERED_BY_REDACTED>
+    started at <REDACTED_TIMESTAMP> (<REDACTED DURATION>s ago) and ran for <REDACTED DURATION>ms
+switch0 failed: failed to look up dendrite clients: proto error: no records found for Query { name: Name("_dendrite._tcp.control-plane.oxide.internal."), query_type: SRV, query_class: IN }
+switch1 failed: failed to look up dendrite clients: proto error: no records found for Query { name: Name("_dendrite._tcp.control-plane.oxide.internal."), query_type: SRV, query_class: IN }
+
 task: "probe_distributor"
   configured period: every <REDACTED_DURATION>m
   last completed activation: <REDACTED ITERATIONS>, triggered by <TRIGGERED_BY_REDACTED>

nexus-config/src/nexus_config.rs

@@ -442,6 +442,8 @@ pub struct BackgroundTaskConfig {
     pub audit_log_timeout_incomplete: AuditLogTimeoutIncompleteConfig,
     /// configuration for audit log cleanup (retention) task
     pub audit_log_cleanup: AuditLogCleanupConfig,
+    /// configuration for populate switch ports task
+    pub populate_switch_ports: PopulateSwitchPortsConfig,
 }
 
 #[serde_as]
@@ -488,6 +490,15 @@ pub struct AuditLogCleanupConfig {
     pub max_deleted_per_activation: u32,
 }
 
+#[serde_as]
+#[derive(Clone, Debug, Deserialize, Eq, PartialEq, Serialize)]
+pub struct PopulateSwitchPortsConfig {
+    /// period (in seconds) for periodic activations of the background task that
+    /// attempts to populate the `switch_port` table.
+    #[serde_as(as = "DurationSeconds<u64>")]
+    pub period_secs: Duration,
+}
+
 #[serde_as]
 #[derive(Clone, Debug, Deserialize, Eq, PartialEq, Serialize)]
 pub struct DnsTasksConfig {
@@ -1339,6 +1350,7 @@ mod test {
             audit_log_cleanup.period_secs = 600
             audit_log_cleanup.retention_days = 90
             audit_log_cleanup.max_deleted_per_activation = 10000
+            populate_switch_ports.period_secs = 31
             [default_region_allocation_strategy]
             type = "random"
             seed = 0
@@ -1620,6 +1632,9 @@ mod test {
                             retention_days: NonZeroU32::new(90).unwrap(),
                             max_deleted_per_activation: 10_000,
                         },
+                        populate_switch_ports: PopulateSwitchPortsConfig {
+                            period_secs: Duration::from_secs(31),
+                        },
                     },
                     multicast: MulticastConfig { enabled: false },
                     default_region_allocation_strategy:
@@ -1735,6 +1750,7 @@ mod test {
             audit_log_cleanup.period_secs = 600
             audit_log_cleanup.retention_days = 90
             audit_log_cleanup.max_deleted_per_activation = 10000
+            populate_switch_ports.period_secs = 31
 
             [default_region_allocation_strategy]
             type = "random"

nexus/Cargo.toml

@@ -118,6 +118,7 @@ slog.workspace = true
 slog-async.workspace = true
 slog-dtrace.workspace = true
 slog-error-chain.workspace = true
+strum.workspace = true
 swrite.workspace = true
 display-error-chain.workspace = true
 slog-term.workspace = true

nexus/background-task-interface/src/init.rs

@@ -62,6 +62,7 @@ pub struct BackgroundTasks {
     pub task_trust_quorum_manager: Activator,
     pub task_attached_subnet_manager: Activator,
     pub task_session_cleanup: Activator,
+    pub task_populate_switch_ports: Activator,
 
     // Handles to activate background tasks that do not get used by Nexus
     // at-large.  These background tasks are implementation details as far as

nexus/examples/config-second.toml

@@ -204,6 +204,7 @@ audit_log_timeout_incomplete.max_timed_out_per_activation = 1000
 audit_log_cleanup.period_secs = 600
 audit_log_cleanup.retention_days = 90
 audit_log_cleanup.max_deleted_per_activation = 10000
+populate_switch_ports.period_secs = 30
 
 [default_region_allocation_strategy]
 # allocate region on 3 random distinct zpools, on 3 random distinct sleds.

nexus/examples/config.toml

@@ -188,6 +188,7 @@ audit_log_timeout_incomplete.max_timed_out_per_activation = 1000
 audit_log_cleanup.period_secs = 600
 audit_log_cleanup.retention_days = 90
 audit_log_cleanup.max_deleted_per_activation = 10000
+populate_switch_ports.period_secs = 30
 
 [default_region_allocation_strategy]
 # allocate region on 3 random distinct zpools, on 3 random distinct sleds.

nexus/src/app/background/init.rs

@@ -145,6 +145,7 @@ use super::tasks::v2p_mappings::V2PManager;
 use super::tasks::vpc_routes;
 use super::tasks::webhook_deliverator;
 use crate::Nexus;
+use crate::app::background::tasks::populate_switch_ports;
 use crate::app::oximeter::PRODUCER_LEASE_DURATION;
 use crate::app::quiesce::NexusQuiesceHandle;
 use crate::app::saga::StartSaga;
@@ -277,6 +278,7 @@ impl BackgroundTasksInitializer {
             task_trust_quorum_manager: Activator::new(),
             task_attached_subnet_manager: Activator::new(),
             task_session_cleanup: Activator::new(),
+            task_populate_switch_ports: Activator::new(),
 
             // Handles to activate background tasks that do not get used by Nexus
             // at-large.  These background tasks are implementation details as far as
@@ -372,6 +374,7 @@ impl BackgroundTasksInitializer {
             task_session_cleanup,
             task_audit_log_timeout_incomplete,
             task_audit_log_cleanup,
+            task_populate_switch_ports,
             // Add new background tasks here.  Be sure to use this binding in a
             // call to `Driver::register()` below.  That's what actually wires
             // up the Activator to the corresponding background task.
@@ -1224,7 +1227,7 @@ impl BackgroundTasksInitializer {
             description: "distributes attached subnets to sleds and switch",
             period: config.attached_subnet_manager.period_secs,
             task_impl: Box::new(attached_subnets::Manager::new(
-                resolver,
+                resolver.clone(),
                 datastore.clone(),
             )),
             opctx: opctx.child(BTreeMap::new()),
@@ -1272,7 +1275,7 @@ impl BackgroundTasksInitializer {
                  than the retention period",
             period: config.audit_log_cleanup.period_secs,
             task_impl: Box::new(audit_log_cleanup::AuditLogCleanup::new(
-                datastore,
+                datastore.clone(),
                 config.audit_log_cleanup.retention_days,
                 config.audit_log_cleanup.max_deleted_per_activation,
             )),
@@ -1281,6 +1284,23 @@ impl BackgroundTasksInitializer {
             activator: task_audit_log_cleanup,
         });
 
+        driver.register(TaskDefinition {
+            name: "populate_switch_ports",
+            description: "one-time population of the `switch_port` table \
+                containing all QSFP ports managed by dendrite",
+            period: config.populate_switch_ports.period_secs,
+            task_impl: Box::new(
+                populate_switch_ports::SwitchPortPopulator::new(
+                    rack_id,
+                    datastore.clone(),
+                    resolver.clone(),
+                ),
+            ),
+            opctx: opctx.child(BTreeMap::new()),
+            watchers: vec![],
+            activator: task_populate_switch_ports,
+        });
+
         driver
     }
 }

nexus/src/app/background/tasks/mod.rs

@@ -37,6 +37,7 @@ pub mod nat_cleanup;
 pub mod networking;
 pub mod phantom_disks;
 pub mod physical_disk_adoption;
+pub mod populate_switch_ports;
 pub mod probe_distributor;
 pub mod read_only_region_replacement_start;
 pub mod reconfigurator_config;