Tag: distributed-systems

All the articles with the tag "distributed-systems".

• The dependency that's 'up' is taking you down

  18 Mar, 2026

  A dead dependency fails fast. A slow one holds threads, exhausts connection pools, and cascades failure to everything upstream. How to treat slow the same as dead with timeouts, bulkheads, and fast-fail.

• FOR UPDATE SKIP LOCKED — and when it stops working

  6 Mar, 2026

  How to calculate the throughput ceiling of SKIP LOCKED, recognize the four degradation signals, and graduate to partition assignment when contention wins.