bongoDev

Loading...

Empowering technology learners worldwide with professional online live trainings. Transform your IT career with expert-led courses.

Facebook Instagram LinkedIn Github

Company

About Us
Careers
Blog
Events

Legal

Privacy Policy
Refund Policy
Terms and Conditions

Contact us

[email protected]
01757258398 01978367094 01576641348
01757258398

Our office

Flat 5/A, Shimanto Plaza, Plot #26/D, Main Road-03, Section #11, Mirpur, Dhaka 1216

© 2026 bongoDev. All rights reserved. | ই-ট্রেড লাইসেন্স: TRADEDNCC/015304/2024

Courses Job Success

Blog Events Free Assessment Login

bongoDev | Event-Driven Reliability Pro - Knowledge Hub

Knowledge Hub/Event-Driven Reliability Pro

Englishবাংলা

Learning path

Event-Driven Reliability Pro

Production reliability engineering for event-driven and messaging systems: delivery semantics, ordering, idempotency, outbox/inbox, retries, DLQs, backpressure, queues, streams, consumer lag, observability, HA/DR, platform semantics, incident response and safe replay.

Discount price

৳ 320.00৳ 120.00Save ৳ 200.00

Lessons

95 published lessons · 1 free preview

01Event-Driven Reliability Mental Model: Publish, Persist, Deliver, Process, AcknowledgeBuild an end-to-end reliability model for asynchronous systems from producer intent to durable business outcome.Beginner · 34 min readFree previewFree preview
02Synchronous vs Asynchronous Architecture: Choose Failure Semantics Deliberately

Decide when asynchronous messaging improves decoupling and when it only moves latency and failure somewhere harder to see.

Beginner · 30 min read

Locked

Locked

03Event, Message, Command, Notification & Task: Model Intent CorrectlySeparate facts, requests and work items so consumers can reason correctly about ownership, replay and side effects.Beginner · 28 min readLockedLocked

04Queues, Topics, Streams & Logs: Pick the Right Delivery ShapeCompare competing-consumer queues, pub/sub topics and retained ordered logs by replay, fan-out, ordering and retention requirements.Beginner · 32 min readLockedLocked

05Producer → Broker → Consumer: Responsibility Handoffs & Failure WindowsIdentify exactly when responsibility for a message transfers and what acknowledgement proves at each boundary.Intermediate · 34 min readLockedLocked

06Delivery Semantics: At-Most-Once, At-Least-Once & Exactly-Once BoundariesReason about delivery guarantees without confusing broker delivery with exactly-once business effects.Intermediate · 36 min readLockedLocked

07Message Envelope Design: IDs, Correlation, Causation, Time, Source & VersionDesign a standard message envelope that supports deduplication, tracing, incident analysis and schema evolution.Intermediate · 38 min readLockedLocked

08Event Naming & Domain Semantics: Facts Should Survive Implementation ChangesCreate domain event names that communicate completed facts instead of leaking transport or database implementation details.Beginner · 28 min readLockedLocked

09Schema Evolution: Backward, Forward & Full CompatibilityEvolve event payloads without breaking independently deployed producers, consumers or replay jobs.Intermediate · 42 min readLockedLocked

10Schema Registry & Contract GovernanceUse versioned schemas, compatibility checks and ownership rules to stop accidental breaking changes from reaching production.Intermediate · 36 min readLockedLocked

11Event Time vs Processing Time: Reason About Late and Delayed EventsSeparate when something happened from when the system observed or processed it.Intermediate · 34 min readLockedLocked

12Ordering Guarantees: Define the Scope Before Depending on OrderDefine per-entity, per-partition, per-key or global ordering explicitly and understand the throughput trade-off.Intermediate · 38 min readLockedLocked

13Partitioning & Sharding: Scale Without Losing Per-Key CorrectnessChoose partition keys that balance throughput while preserving the ordering relationships the domain actually needs.Intermediate · 40 min readLockedLocked

14Hot Partitions, Hot Keys & SkewDetect uneven traffic that makes average utilization hide a saturated partition or consumer.Advanced · 38 min readLockedLocked

15Retention, Replay & Event HistoryChoose retention and replay windows based on recovery, audit, reprocessing and cost requirements.Intermediate · 36 min readLockedLocked

16Consumer Groups & Competing ConsumersScale consumers while understanding partition ownership, work sharing, and the difference between fan-out and load sharing.Intermediate · 34 min readLockedLocked

17Acknowledgements, Visibility Timeouts & Processing LeasesTreat acknowledgements and visibility/lease deadlines as ownership protocols tied to successful processing.Intermediate · 40 min readLockedLocked

18Acknowledge After the Durable Side EffectPlace acknowledgements after the state transition that must survive a consumer crash.Intermediate · 34 min readLockedLocked

19Producer Confirms & Publish DurabilityUse broker acknowledgements, confirms and idempotent publisher logic instead of assuming a successful socket write equals durable publish.Intermediate · 38 min readLockedLocked

20Ambiguous Publish Outcomes & Producer DeduplicationDesign producers so uncertain publish results do not become duplicate business intent or silent loss.Advanced · 38 min readLockedLocked

21Idempotent Consumers: Make Redelivery SafeMake repeated delivery produce one correct business outcome rather than repeated side effects.Intermediate · 42 min readLockedLocked

22Deduplication Stores, Unique Constraints & Retention WindowsChoose deduplication keys and retention windows that match redelivery and replay behavior.Advanced · 36 min readLockedLocked

23Transactional Outbox: Database State and Event Publication as One IntentAvoid dual-write gaps by committing business state and an outbox record in one local database transaction.Advanced · 46 min readLockedLocked

24Inbox Pattern: Atomic Consumer Deduplication & State ChangeUse an inbox/processed-message record with the consumer state mutation so duplicate delivery can be safely acknowledged.Advanced · 42 min readLockedLocked

25Change Data Capture (CDC): Reliability Between Database Log and Event StreamUnderstand CDC offset/checkpoint, schema and replay failure modes when database changes become event streams.Advanced · 42 min readLockedLocked

26Saga Pattern: Distributed Business Transactions Without a Global LockCoordinate multi-service workflows with explicit state, compensations and failure handling.Advanced · 46 min readLockedLocked

27Orchestration vs ChoreographyChoose explicit workflow coordination or event choreography based on complexity, visibility, coupling and recovery needs.Advanced · 40 min readLockedLocked

28Exactly-Once Processing: Define the Transactional BoundaryDistinguish broker-level processing guarantees from external side effects, databases, APIs and human actions.Advanced · 46 min readLockedLocked

29Retry Classification: Transient, Persistent & Business FailuresRetry only failures that have a realistic chance of succeeding and keep invalid business input out of retry loops.Intermediate · 34 min readLockedLocked

30Exponential Backoff & Jitter for ConsumersSpread retry pressure over time so a dependency recovery does not trigger a synchronized retry storm.Intermediate · 34 min readLockedLocked

31Retry Topics / Retry Queues: Keep Failure Delays Out of Hot ConsumersUse staged retry paths when immediate requeue causes tight loops and consumer starvation.Intermediate · 38 min readLockedLocked

32Poison Messages: Detect Deterministic Consumer FailureRecognize messages that will never succeed without code, data or configuration change.Intermediate · 36 min readLockedLocked

33Dead-Letter Queues: Quarantine, Not a GraveyardDesign DLQs with ownership, observability, retention and a safe repair/redrive path.Intermediate · 38 min readLockedLocked

34Safe DLQ Redrive: Repair Before ReplayPrevent mass redrive from recreating the original outage or duplicating downstream side effects.Advanced · 40 min readLockedLocked

35Message TTL, Expiration & Delayed WorkUse TTL and delayed scheduling without silently dropping work that is still business-critical.Intermediate · 32 min readLockedLocked

36Backpressure: Let Downstream Capacity Control Upstream WorkStop consumers and producers from creating unbounded in-flight work when a dependency slows down.Advanced · 40 min readLockedLocked

37Prefetch / Max In-Flight: Throughput vs Fairness vs RecoveryTune in-flight messages so consumers stay busy without hoarding work or multiplying failure blast radius.Advanced · 38 min readLockedLocked

38Queue Depth vs Consumer Lag vs Oldest Message AgeUse backlog measurements that reflect user waiting time rather than relying on one raw message count.Intermediate · 36 min readLockedLocked

39Drain Time & Recovery CapacityCalculate whether consumers can catch up after an outage instead of only matching current arrival rate.Advanced · 38 min readLockedLocked

40Event-Driven Autoscaling: Scale on Work, Not Only CPUScale consumers using lag, queue depth, oldest age or event rate while protecting downstream capacity.Advanced · 42 min readLockedLocked

41Overload Protection: Admission Control, Quotas & Load SheddingProtect critical event flows when producer rate exceeds sustainable system capacity.Advanced · 38 min readLockedLocked

42Head-of-Line BlockingPrevent one slow or blocked message sequence from delaying unrelated work.Advanced · 36 min readLockedLocked

43Slow Consumers & Consumer StarvationDetect consumers that are alive but not making useful progress.Intermediate · 34 min readLockedLocked

44Consumer Rebalances, Ownership Changes & Duplicate WindowsUnderstand how membership changes can pause work and create reprocessing around ownership handoff.Advanced · 40 min readLockedLocked

45Request/Reply over Messaging: Correlation, Timeouts & OrphansUse asynchronous request/reply only with explicit correlation, timeout, late-response and cleanup behavior.Advanced · 38 min readLockedLocked

46Publish/Subscribe Fan-Out & Independent Consumer FailureDesign fan-out so a broken analytics subscriber cannot block billing or core business consumers.Intermediate · 34 min readLockedLocked

47Work Queue / Competing Consumer PatternDistribute independent jobs across workers while preserving idempotency and fair retry behavior.Intermediate · 34 min readLockedLocked

48Event Log / Stream Pattern: Replayable State ChangesUse retained streams when consumers need independent positions, replay and derived-state rebuilding.Intermediate · 34 min readLockedLocked

49Kafka Reliability Integration: What This Track Adds Beyond Kafka ProApply cross-system reliability patterns to Kafka without repeating broker internals taught in Kafka Pro.Advanced · 40 min readLockedLocked

50RabbitMQ Reliability: Publisher Confirms, Manual Acks, Quorum Queues & DLXOperate RabbitMQ with explicit producer confirms, consumer acknowledgements and durable queue topology choices.Advanced · 44 min readLockedLocked

51Amazon SQS Reliability: Visibility, Long Polling, DLQ & FIFO Message GroupsDesign SQS consumers around visibility timeout, deletion, redelivery, dead-lettering and ordering scope.Advanced · 42 min readLockedLocked

52Google Cloud Pub/Sub Reliability: Ack Deadlines, Ordering & Exactly-Once ScopeDesign Pub/Sub subscriptions with acknowledgement deadlines, redelivery, ordering keys and explicit exactly-once boundaries.Advanced · 42 min readLockedLocked

53Azure Service Bus Reliability: Peek-Lock, Sessions, DLQ & Duplicate DetectionApply reliable receive-lock, settlement, ordered-session and dead-letter patterns in Azure Service Bus.Advanced · 42 min readLockedLocked

54NATS JetStream Reliability: Streams, Consumers, Acks & RedeliveryUnderstand retained NATS JetStream consumption, acknowledgement and replay trade-offs for low-latency event systems.Advanced · 38 min readLockedLocked

55Choose the Messaging Platform by Semantics, Not PopularityCompare log, queue and pub/sub systems against ordering, replay, latency, durability, throughput and operations requirements.Intermediate · 38 min readLockedLocked

56Multi-Tenancy, Quotas & Noisy-Neighbor IsolationPrevent one tenant or producer from consuming shared broker, partition, subscription or consumer capacity.Advanced · 40 min readLockedLocked

57Authentication, Authorization, TLS & Least PrivilegeTreat event infrastructure as production data plane with scoped producer, consumer and administrative identities.Intermediate · 38 min readLockedLocked

58Sensitive Data, PII & Event Payload MinimizationKeep secrets and unnecessary personal data out of long-lived, replayable and widely consumed event streams.Intermediate · 38 min readLockedLocked

59Auditability & Event LineagePreserve enough identity and causation metadata to reconstruct who produced, transformed and consumed critical events.Advanced · 36 min readLockedLocked

60Event Observability: Producer, Broker, Consumer & Business SignalsMonitor the complete async path rather than declaring health because brokers are reachable.Intermediate · 42 min readLockedLocked

61Distributed Tracing Across Asynchronous BoundariesPropagate trace context through messages while preserving event identity and asynchronous causality.Advanced · 42 min readLockedLocked

62OpenTelemetry Messaging Spans, Links & Async CausalityUse producer/consumer spans and span links when asynchronous processing does not fit a simple parent-child lifetime.Advanced · 42 min readLockedLocked

63Structured Logging & Correlation for Event PipelinesLog message identity, attempt, partition/subscription and outcome without leaking sensitive payloads.Intermediate · 36 min readLockedLocked

64SLOs for Asynchronous Systems: Freshness, Completion & LossDefine reliability objectives that measure time-to-business-outcome, not only broker uptime.Advanced · 42 min readLockedLocked

65Alerting on Backlog: Page on Impact, Not Every QueueCreate alerts based on sustained queue age, drain risk and user impact instead of static queue-depth thresholds.Advanced · 38 min readLockedLocked

66Capacity Planning for Event PipelinesPlan producer, broker, network, storage and consumer capacity as one throughput system.Advanced · 42 min readLockedLocked

67Backlog Storage, Retention & Disk RiskEstimate storage growth during downstream outages and prevent retention from becoming an emergency limit.Advanced · 36 min readLockedLocked

68High Availability & Broker Failure DomainsSeparate broker availability from client correctness, replica quorum, consumer recovery and downstream redundancy.Advanced · 40 min readLockedLocked

69Multi-Region Event Architecture: Active/Passive, Active/Active & Conflict RiskChoose cross-region event patterns based on RPO/RTO, ordering, sovereignty, duplication and reconciliation constraints.Advanced · 44 min readLockedLocked

70Disaster Recovery for Messaging: Restore the Pipeline, Not Just the BrokerDesign recovery for topics/queues, schemas, ACLs, offsets, consumer state, replay and downstream dependencies.Advanced · 42 min readLockedLocked

71Event Replay Safety: Rebuild Derived State Without Repeating External EffectsSeparate replayable projection logic from non-repeatable side effects such as email, payment and partner calls.Advanced · 42 min readLockedLocked

72Backfills & Historical ReprocessingRun large historical jobs without starving live traffic or violating ordering and capacity assumptions.Advanced · 40 min readLockedLocked

73Stateful Stream Processing: Checkpoints, State Stores & RecoveryUnderstand that stream processors have data state plus processing position that must recover consistently.Advanced · 44 min readLockedLocked

74Windowing, Watermarks & Late EventsHandle out-of-order and late data without pretending processing arrival order equals business time.Advanced · 44 min readLockedLocked

75Release Engineering for Producers & ConsumersDeploy event producers and consumers independently while maintaining contract compatibility and rollback safety.Advanced · 40 min readLockedLocked

76Consumer Version Skew & Rolling DeploymentsExpect multiple consumer versions to process the same stream during rolling releases and rebalances.Advanced · 38 min readLockedLocked

77Configuration & Secret Changes in Event SystemsTreat broker endpoints, credentials, TLS certificates, topic names and retry settings as production releases.Intermediate · 38 min readLockedLocked

78Testing Event-Driven Systems: Contracts, Integration, Replay & FailureBuild tests that exercise duplicate delivery, reordering, redelivery, delayed events and dependency failure—not only happy-path payloads.Intermediate · 42 min readLockedLocked

79Deterministic Consumer Tests with Duplicate & Out-of-Order EventsMake correctness properties executable by feeding the same event multiple times and in adversarial order.Advanced · 40 min readLockedLocked

80Load & Soak Testing Event PipelinesMeasure sustainable publish/consume rates, tail latency, backlog growth, recovery and long-running leak behavior.Advanced · 42 min readLockedLocked

81Chaos & Resilience Testing for MessagingInject broker, network, consumer and dependency failures with guardrails to verify redelivery and recovery assumptions.Advanced · 42 min readLockedLocked

82Event-Driven Incident Patterns: Lost, Duplicated, Delayed, Reordered & StuckClassify async incidents by observable symptom before changing broker or consumer configuration.Advanced · 44 min readLockedLocked

83Incident: Queue Backlog & Consumer LagTriage backlog by arrival rate, processing capacity, oldest age, partition skew and downstream saturation.Advanced · 42 min readLockedLocked

84Incident: Duplicate Business Side EffectsTrace duplicate effects back through publish retry, redelivery, timeout ambiguity and missing idempotency.Advanced · 42 min readLockedLocked

85Incident: Poison Message & Retry StormStop deterministic failures from consuming all worker capacity while preserving evidence and repairability.Advanced · 40 min readLockedLocked

86Incident: Broker Healthy, Business Pipeline BrokenTroubleshoot when infrastructure health is green but the asynchronous customer journey is not completing.Advanced · 40 min readLockedLocked

87Lab: Reliable Order Pipeline with Outbox, Idempotent Consumer & DLQBuild an order event pipeline that survives producer crash, duplicate delivery and poison messages without duplicate business effects.Advanced · 60 min readLockedLocked

88Lab: Retry Storm, Backpressure & Controlled RecoveryCreate a downstream slowdown, observe retry amplification, then recover using bounded retries, jitter and concurrency limits.Advanced · 60 min readLockedLocked

89Lab: Poison Message, DLQ Repair & Canary RedriveOperate the full poison-message lifecycle from detection through quarantine, code repair and rate-limited replay.Advanced · 55 min readLockedLocked

90Lab: Consumer Lag, Autoscaling & Downstream Capacity GuardrailScale workers from backlog while enforcing the maximum concurrency the database or downstream API can sustain.Advanced · 60 min readLockedLocked

91Lab: Kafka/RabbitMQ/SQS Semantics ComparisonImplement the same idempotent job workflow on different messaging primitives and compare ack, retry, ordering and replay semantics.Advanced · 60 min readLockedLocked

92Lab: Regional Failure & Replay-Based RecoverySimulate regional messaging failure, recover service, replay safely and reconcile any duplicate or stale derived state.Advanced · 65 min readLockedLocked

93Event-Driven Reliability Troubleshooting PlaybookTroubleshoot event pipelines systematically from business symptom to producer, broker, consumer, dependency and acknowledgement boundary.Advanced · 58 min readLockedLocked

94Event-Driven Command & Evidence PlaybookTurn Kafka, RabbitMQ, SQS, Pub/Sub and application commands into explicit questions, evidence and limitations.Advanced · 58 min readLockedLocked

95Event-Driven Reliability Pro CheatsheetConsolidate delivery semantics, ordering, idempotency, retries, DLQ, backpressure, observability and recovery into one production reference.Advanced · 55 min readLockedLocked