AgentRules.org
RulesStacksClientsAuditorPrunerSelf-HealRFCsBenchmarkCLICI/CDAuthorsSubmit
This specification is maintained by the AgentRules Working Group. Have enhancements or want to claim maintainership?Submit Specification →
devopscritical PriorityID: otel-arrow-memory-pool-fencing

OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing & Heap Quotas

Enforces strict heap allocation quotas and bounded memory pool reservation across Apache Arrow record batches within OpenTelemetry collectors. Prevents out-of-memory (OOM) fatal cascades during high-cardinality telemetry storms by requiring custom allocator fencing, backpressure rejection, and deterministic off-heap buffer disposal.

Prohibited Anti-Pattern: Unbounded Direct Heap Allocation of Arrow Record Batches Causing Fatal Collector OOM Under Metric Ingestion Spikes

Export Ready-to-Drop Configuration

Select format below to copy code or download file directly.

---
description: OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing & Heap Quotas - Enforces strict heap allocation quotas and bounded memory pool reservation across Apache Arrow record batches within OpenTelemetry collectors. Prevents out-of-memory (OOM) fatal cascades during high-cardinality telemetry storms by requiring custom allocator fencing, backpressure rejection, and deterministic off-heap buffer disposal.
globs: **/collector/**/*.go, **/otlp/**/*.go, **/arrow/**/*.go, **/pipeline/**/*.go
alwaysApply: true
verified_by: "agentrules.org"
license: "CC-BY-4.0"
source: "https://agentrules.org/rules/otel-arrow-memory-pool-fencing"
---

# OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing & Heap Quotas

> **Severity**: CRITICAL | **Category**: devops
> **Anti-Pattern Warning**: Unbounded Direct Heap Allocation of Arrow Record Batches Causing Fatal Collector OOM Under Metric Ingestion Spikes

## System Constraints
You are a Principal Engineer enforcing OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing. Mandate strict memory pool boundaries, custom Arrow memory allocators, explicit batch size ceilings, deterministic memory recycling via Release(), and early circuit-breaking backpressure before the Go runtime or off-heap allocations trigger SIGKILL.

## Strict Guidelines
- Engineers MUST ALWAYS instantiate Apache Arrow record builders using a bounded `memory.Allocator` wrapper that tracks cumulative bytes and enforces a hard ceiling.
- Engineers MUST NOT rely on Go's default GC or unmetered `memory.DefaultAllocator` for Arrow columnar buffers handling untrusted external telemetry payloads.
- Pipelines MUST immediately drop or reject ingestion batches with HTTP 429/gRPC `ResourceExhausted` when the fenced memory pool reaches the pre-configured watermark (e.g., 80% soft limit, 90% hard limit).
- Services MUST invoke `Record.Release()` deterministically in `defer` statements or downstream sink handoffs to guarantee immediate buffer deallocation back to the memory pool.
- Components NEVER pass uncompressed columnar Arrow record batches downstream without verifying that downstream consumer allocation capacity is pre-reserved.

## Bad vs Good Implementations
### Bounded Memory Pool Allocation with Immediate Release

#### Bad Antipattern (AI Common Hallucination)
> **Why it fails**: Using memory.DefaultAllocator allocates unmetered heap memory without backpressure. When cardinality spikes, Arrow builders double allocations dynamically, exhausting physical RAM and triggering the OS OOM killer before GC runs.
```go
package collector

import (
	"context"
	"github.com/apache/arrow/go/v17/arrow/array"
	"github.com/apache/arrow/go/v17/arrow/memory"
	"go.opentelemetry.io/collector/pdata/pmetric"
)

func IngestMetrics(ctx context.Context, metrics pmetric.Metrics) error {
	// ANTI-PATTERN: Default unmetered allocator without capacity ceiling or release
	pool := memory.DefaultAllocator
	b := array.NewRecordBuilder(pool, MetricSchema)
	
	// If metrics storm arrives, heap expands uncontrollably leading to OOM-killer SIGKILL
	for i := 0; i < metrics.ResourceMetrics().Len(); i++ {
		b.Field(0).(*array.Float64Builder).Append(metrics.ResourceMetrics().At(i).SchemaUrl())
	}
	record := b.NewRecord()
	return PushDownstream(record)
}
```

#### Production Standard (Correct)
> **Standard**: Uses a fenced allocator that monitors current allocations against a strict quota. When the quota is exceeded, it immediately rejects incoming payloads with gRPC ResourceExhausted (429), protecting the process. Both builder and records are deterministically released to return memory to the pool.
```go
package collector

import (
	"context"
	"errors"
	"github.com/apache/arrow/go/v17/arrow/array"
	"github.com/apache/arrow/go/v17/arrow/memory"
	"go.opentelemetry.io/collector/pdata/pmetric"
	"google.golang.org/grpc/codes"
	"google.golang.org/grpc/status"
)

var ErrQuotaExceeded = status.Error(codes.ResourceExhausted, "arrow memory pool quota exceeded")

type FencedPool struct {
	*memory.CheckedAllocator
	maxQuota int64
}

func IngestMetrics(ctx context.Context, metrics pmetric.Metrics, pool *FencedPool) error {
	// PRODUCTION INVARIANT: Bounded checked allocator with backpressure and deterministic release
	if pool.CurrentAlloc() > pool.maxQuota {
		return ErrQuotaExceeded
	}

	b := array.NewRecordBuilder(pool, MetricSchema)
	defer b.Release()

	for i := 0; i < metrics.ResourceMetrics().Len(); i++ {
		b.Field(0).(*array.StringBuilder).Append(metrics.ResourceMetrics().At(i).SchemaUrl())
	}
	record := b.NewRecord()
	defer record.Release()

	return PushDownstream(record.NewSlice(0, record.NumRows()))
}
```
Recommended MCP Servers (mcp-serverlist.com)

Equip agent with system tools to verify code for this stack.

Prometheus Collector Monitoring
Monitors otelcol_process_memory_rss and arrow_memory_pool_allocated_bytes to trigger alert thresholds during telemetry storms.
Recommended Agent Skills (askills.dev)

Ready-to-use executable skill functions tailored to this stack.

OpenTelemetry Benchmarking Suite
Executes high-cardinality synthetic metric storms to validate memory allocator fencing and backpressure mechanisms.
Architectural Specification & Invariant Rationale

Architectural Invariant Design: OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing & Heap Quotas

OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing & Heap Quotas defines an industrial-grade behavioral standard for devops engineering. Enforces strict heap allocation quotas and bounded memory pool reservation across Apache Arrow record batches within OpenTelemetry collectors. Prevents out-of-memory (OOM) fatal cascades during high-cardinality telemetry storms by requiring custom allocator fencing, backpressure rejection, and deterministic off-heap buffer disposal. Modern software architectures require strict adherence to deterministic patterns to guarantee high availability, maintainability, and zero security vulnerabilities.

Framework Paradigm Shift & Drift MechanicsLLMs are trained on millions of open-source repositories containing outdated code, legacy idioms, and unpatched antipatterns. When coding assistants generate implementations for this stack, they default to common statistical tokens: Unbounded Direct Heap Allocation of Arrow Record Batches Causing Fatal Collector OOM Under Metric Ingestion Spikes Without an explicit system instruction file, developers spend hours debugging silent failures.
Failure Modes & Production Blast Radius

Hazard Analysis: What Breaks Without This Rule

The following failure modes have been cataloged from hundreds of hours of production code reviews where AI assistants produced non-compliant implementations:

Failure ModeAI Symptom (The Anti-Pattern)SeverityProduction Impact & Remedy
AI Hallucination & Deprecation DriftUnbounded Direct Heap Allocation of Arrow Record Batches Causing Fatal Collector OOM Under Metric Ingestion SpikesCritical
Violating this industrial standard introduces regression bugs, security risks, and unexpected runtime failures in production.
Enforce explicit behavioral system directives across all AI coding assistants in your repository.
Silent Architecture InconsistencyAI assistants produce mixed paradigms across files, creating fragmented codebase conventions.Medium
Increases cognitive load during code reviews and leads to maintenance debt across engineering teams.
Pin project conventions in .cursor/rules, CLAUDE.md, and AGENTS.md so all team agents follow identical guardrails.
Architectural Directives in Detail

Step-by-Step Production Directives

Each rule constraint below is injected into the AI agent's system prompt to establish deterministic boundaries:

1

Engineers MUST ALWAYS instantiate Apache Arrow record builders using a bounded `memory.Allocator` wrapper that tracks cumulative bytes and enforces a hard ceiling.

Engineers MUST ALWAYS instantiate Apache Arrow record builders using a bounded `memory.Allocator` wrapper that tracks cumulative bytes and enforces a hard ceiling.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.
Enforcement:Integrated into the Agent Pack rule file. When Cursor, Claude Code, or Codex processes this prompt, it applies this constraint before emitting syntax.
2

Engineers MUST NOT rely on Go's default GC or unmetered `memory.DefaultAllocator` for Arrow columnar buffers handling untrusted external telemetry payloads.

Engineers MUST NOT rely on Go's default GC or unmetered `memory.DefaultAllocator` for Arrow columnar buffers handling untrusted external telemetry payloads.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.
Enforcement:Integrated into the Agent Pack rule file. When Cursor, Claude Code, or Codex processes this prompt, it applies this constraint before emitting syntax.
3

Pipelines MUST immediately drop or reject ingestion batches with HTTP 429/gRPC `ResourceExhausted` when the fenced memory pool reaches the pre-configured watermark (e.g., 80% soft limit, 90% hard limit).

Pipelines MUST immediately drop or reject ingestion batches with HTTP 429/gRPC `ResourceExhausted` when the fenced memory pool reaches the pre-configured watermark (e.g., 80% soft limit, 90% hard limit).

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.
Enforcement:Integrated into the Agent Pack rule file. When Cursor, Claude Code, or Codex processes this prompt, it applies this constraint before emitting syntax.
4

Services MUST invoke `Record.Release()` deterministically in `defer` statements or downstream sink handoffs to guarantee immediate buffer deallocation back to the memory pool.

Services MUST invoke `Record.Release()` deterministically in `defer` statements or downstream sink handoffs to guarantee immediate buffer deallocation back to the memory pool.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.
Enforcement:Integrated into the Agent Pack rule file. When Cursor, Claude Code, or Codex processes this prompt, it applies this constraint before emitting syntax.
5

Components NEVER pass uncompressed columnar Arrow record batches downstream without verifying that downstream consumer allocation capacity is pre-reserved.

Components NEVER pass uncompressed columnar Arrow record batches downstream without verifying that downstream consumer allocation capacity is pre-reserved.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.
Enforcement:Integrated into the Agent Pack rule file. When Cursor, Claude Code, or Codex processes this prompt, it applies this constraint before emitting syntax.
Technical Questions & Answers

Frequently Asked Technical Questions

Common questions developers face when configuring AI coding assistants for this stack:

Q:Why is OpenTelemetry OTLP/Arrow Columnar Memory Pool Fencing & Heap Quotas critical for AI coding assistants?

Enforces strict heap allocation quotas and bounded memory pool reservation across Apache Arrow record batches within OpenTelemetry collectors. Prevents out-of-memory (OOM) fatal cascades during high-cardinality telemetry storms by requiring custom allocator fencing, backpressure rejection, and deterministic off-heap buffer disposal. Without explicit behavioral boundaries, AI coding assistants (Cursor, Claude Code, Windsurf, Codex) rely on training data averages, which frequently produce deprecated, inefficient, or vulnerable code patterns.

Q:How does adopting this rule prevent AI hallucinations?

This rule explicitly declares what the AI must avoid (Unbounded Direct Heap Allocation of Arrow Record Batches Causing Fatal Collector OOM Under Metric Ingestion Spikes) and prescribes verified positive examples. Grounding the LLM with negative and positive pairs eliminates guesswork.

Q:Which AI coding environments are supported?

This specification is multi-platform compatible: Cursor (.cursor/rules/*.mdc & .cursorrules), Anthropic Claude Code (CLAUDE.md), Windsurf (.windsurfrules), Codex / Cline (AGENTS.md), and GitHub Copilot (.github/copilot-instructions.md).

Q:Can I combine this rule with other technology stacks?

Yes. Use the Interactive Stack Builder or download an Agent Pack to merge this rule with other frontend, backend, database, security, and DevOps rules without syntax collision.

Open Specification · CC BY 4.0·Free for Personal, Team & Commercial In-Repo Use

Citation & Open Source Attribution Guidelines

This specification is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). You are completely free to adapt and share this rule in your workspace. If citing, including in public AI benchmark datasets, or redistributing, attribution to AgentRules.org is required.

Related DEVOPS Rules

View all rules →
devops
86

Kubernetes Gateway API v1 HTTPRoute Canary and ReferenceGrant Invariants

Enforces strict conformance to Kubernetes Gateway API v1 specifications during migration from legacy Ingress controllers. Mandates explicit ReferenceGrant resources for cross-namespace routing, strict relative integer weighting for HTTPRoute canary releases, and explicit backendRef port declarations.

devops
118

ArgoCD ApplicationSet Matrix Generator Invariants & Cascade Deletion Guards

Enforces strict structural constraints on ArgoCD ApplicationSet Matrix generators combining cluster selectors and Git directories. Prevents catastrophic multi-cluster cascade pruning, accidental secret exposure through parameter interpolation, and runaway Application CR explosion via explicit sync policies, preservation finalizers, and parameter validation.

devops
40

Karpenter Resilient Topology, Interruption Queue, and Consolidation Budget Standard

Enforces deterministic disruption budgets, multi-instance-type diversification, and event-driven SQS interruption queue handling across Karpenter NodePools and EC2NodeClasses. Eliminates cascading eviction loops, instance starvation, and availability downtime during AWS Spot rebalances and underutilized node consolidation.

1 Rule in Stack

OpenTelemetry