trigger_manual_compaction No compactor is available. compact

Error message

trigger_manual_compaction No compactor is available. compaction_group {}

What it means

trigger_manual_compaction requires an idle compactor to run the requested compaction task. When compactor_manager.next_compactor() returns None — no compactor is connected or idle — the request fails with this error naming the compaction group.

Source

Thrown at src/meta/src/hummock/manager/compaction/mod.rs:1088

            self.try_send_compaction_request(compaction_group, compact_task::TaskType::Dynamic);
        }
        Ok(())
    }

    pub async fn trigger_manual_compaction(
        &self,
        compaction_group: CompactionGroupId,
        manual_compaction_option: ManualCompactionOption,
    ) -> Result<ManualCompactionTriggerResult> {
        let start_time = Instant::now();
        let exclusive = manual_compaction_option.exclusive;

        // 1. Get idle compactor.
        let compactor = match self.compactor_manager.next_compactor() {
            Some(compactor) => compactor,
            None => {
                tracing::warn!("trigger_manual_compaction No compactor is available.");
                return Err(anyhow::anyhow!(
                    "trigger_manual_compaction No compactor is available. compaction_group {}",
                    compaction_group
                )
                .into());
            }
        };

        // 2. Get manual compaction task.
        let compact_task = self
            .manual_get_compact_task_with_info(compaction_group, manual_compaction_option)
            .await;
        let (compact_task, blocked_by_pending) = match compact_task {
            Ok((compact_task, blocked_by_pending)) => (compact_task, blocked_by_pending),
            Err(err) => {
                tracing::warn!(error = %err.as_report(), "Failed to get compaction task");
                if matches!(err, Error::InvalidManualCompactionOption(_)) {
                    return Err(err);
                }

View on GitHub (pinned to 6469eb736d)

Solutions

Ensure compactor processes are running and registered (`risectl hummock list-compactor` / check meta logs for compactor registration)
Scale up compactor replicas or wait for running tasks to finish, then retry
Check compactor-to-meta connectivity and advertise address configuration
Retry the manual compaction after confirming an idle compactor is available

Example fix

// before
risectl hummock manual-compaction --group 1   // fails: no compactor
// after
risedev d   # or deploy a compactor so it registers with meta
risectl hummock manual-compaction --group 1

Defensive patterns

Strategy: retry

Validate before calling

// Check compactor availability before issuing manual compaction
let compactors = risectl::hummock::list_compactors()?;
if compactors.is_empty() {
    return Err(anyhow!("no compactor registered; start compactor first"));
}

Type guard

fn compactor_available(cm: &CompactorManager) -> bool { cm.next_compactor().is_some() }

Try / catch

match trigger_manual_compaction(group).await {
    Err(e) if e.to_string().contains("No compactor is available") => {
        tracing::warn!("no idle compactor; retrying in 10s");
        tokio::time::sleep(Duration::from_secs(10));
        // retry with backoff
    }
    other => other?,
}

Prevention

Run enough compactor replicas and alert on zero registrations
Check compactor→meta connectivity and advertise addresses
Schedule manual compactions away from heavy automatic compaction windows
Add retry-with-backoff around manual compaction triggers

When it happens

Trigger: Manual compaction issued (risectl/RPC) while no compactor workers are registered with the meta node, or all registered compactors are busy so next_compactor() has none available.

Common situations: Compactor deployment missing or crashed in the cluster; compactor unable to register (network/config); all compactors saturated by automatic compaction when a manual one is requested; scaling down compactor replicas to zero.

Understand the failure class

Background: 'Could not be found', 'does not exist', 'not found in database': the resource-not-found family when an ID, slug, key, or URI lookup comes back empty — this error's family across 20 libraries.

Related errors

AI-assisted analysis of risingwavelabs/risingwave@6469eb736d (2026-09-11). Data as JSON: /api/errors/250de6e1dba74f41. Report an issue: GitHub.