Arkanthos
Field notes on building and running reliable internet services.
When to Choose a Queue Over a Request
May 31, 2026
Direct synchronous communication remains standard practice because execution flow is intuitive: an upstream service queries, the target responds, and delay is immediately measurable. Message queues become necessary when processing duration exceeds HTTP timeouts or when traffic spikes must be leveled to protect datastores.
Introducing message brokers transforms transient failures into accumulated backlog. While an unhandled exception in an RPC call fails a solitary transaction, an unprocessable payload in a queue can stall an entire consumer group, making dead-letter routing an indispensable launch requirement.
The Operator's Guide to Load Testing
August 1, 2026
Load tests fail to predict production for a consistent reason: the traffic shape is wrong. Uniform random requests against one endpoint tell you about that endpoint. The production killer is the burst of clients retrying in sync after a thirty-second blip, or the cache-cold crawl at 5 a.m. when the …
What Good Observability Actually Looks Like
August 22, 2026
Monitoring consoles sprawl uncontrollably while offering little insight during live incidents. True observability operates under inverted priorities: an on-call engineer gets paged, and telemetry systems must identify the root diff within sixty seconds.…
Email Deliverability: Authentication Beyond SPF
June 4, 2026
Relying strictly on SPF validation is long obsolete. Outbound mail can easily pass envelope checks while failing domain alignment, prompting modern receiving systems to evaluate cryptographic DKIM signatures and strict DMARC enforcement instead.…
Cache Invalidation Patterns That Survive Traffic
April 20, 2026
Engineers constantly repeat the classic computer science joke about cache invalidation, yet few design caches resilient to thundering herds. Request coalescing remains the critical design pattern: upon key expiration, an isolated worker fetches fresh data while concurrent requests receive stale reco…
More reading
- A Practical Guide to API Rate Limiting — Engineering, June 3, 2026
- Multi-Region Failover Planning — Operations, April 24, 2026
- A Field Guide to Graceful Degradation — Engineering, July 5, 2026
About us
We cover the unglamorous middle of software: queues that back up, caches that lie, and DNS at 3 a.m. Everything is written from real operational experience.