Release 1.71.7
Release Notes - Seatmap Platform
Version 1.71.7 - 2026-08-25
Release Focus: The converter keeps answering health checks and metrics while it is busy converting, and accepts a larger volume of conversion work. Delivered as a patch on top of the 1.71 line.
What’s Fixed
The converter reports its status while it is busy converting
The converter applies a request budget to its conversion endpoints. Its health and metrics endpoints now sit outside that budget, so a converter that is working through a queue of conversions continues to answer an orchestrator’s health probe and a metrics collector.
Previously those probes drew on the same budget as conversion traffic. On a busy instance the budget could be spent before a probe arrived, and the probe was refused — so a converter that was running normally and serving requests was reported unhealthy. Where a deployment holds other services back until the converter reports healthy, nothing else started.
The budget for the conversion endpoints themselves is also larger, so a bulk background regeneration started from the editor runs to completion instead of being turned away partway through.
Issue: SEAT-1213.
Affected products
converter-service— health and metrics endpoints excluded from the request budget, larger default budget, and both the budget and its window made configurable.
Upgrade
No API or database changes, and no action is required. The new defaults apply on upgrade.
The budget is configurable if you want to tune it. With the Helm chart:
converter:
config:
rateLimit:
max: 600
windowMs: 60000
Deployments that run the container directly can set RATE_LIMIT_MAX and RATE_LIMIT_WINDOW_MS instead. max is the number of requests allowed per client within the window, and applies only to the conversion endpoints. Setting it to 0 refuses every request to those endpoints.