From 30e97c45d9534cde2dd126d4648ce6c7829fe1f3 Mon Sep 17 00:00:00 2001 From: Pablo Vilas Date: Wed, 2 Sep 2026 15:36:22 -0300 Subject: [PATCH 1/3] docs(changelog): tracing for k8s scope and deployment workflows (#215) --- CHANGELOG.md | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/CHANGELOG.md b/CHANGELOG.md index 3486338d..eee43447 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -6,6 +6,11 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [Unreleased] +- k8s scopes now report their lifecycle and deployment operations to the nullplatform tracing API, so the dashboard shows the same step-by-step timeline for these scopes as for native ones: create, update and delete of the scope, and every deployment action (initial, blue/green, switch traffic, finalize, rollback, delete, diagnose, kill instance, restart pods, pause/resume autoscaling, set desired instance count). Each step shows its duration, long waits report live what they are waiting on (load balancer, DNS, instance health), and each workflow and step carries a human-readable title. Requires `NP_API_KEY` on the agent; without it, or without the bundled tracing SDK, behaviour is unchanged +- k8s scope deployments that fail now name the cause on the step where it happened, with what to do about it: image pull errors with the registry's message, out-of-memory kills, crash exit codes with the application's last log lines, failing startup/readiness probes with the path and response detected. The same message reaches the workflow engine's failure report instead of a bare exit status +- k8s **diagnose** checks now state their finding on their own step (summary, severity, affected pods, recommendation), so the deployment page shows what a check found instead of a list of green steps +- k8s scope and deployment workflows declare a per-provider job identity (`entity`, `operation`, `scope.provider`), so the dashboard can show the pending checklist for an operation before the agent picks the command up +- k8s scope deployments record lineage on the trace: the namespace, deployment, service and ingress they applied or removed, the DNS record and load balancer they depend on, and the build asset they deployed - k8s scopes can now pin the traffic-manager sidecar version cluster-wide via the container-orchestration provider's `traffic_manager.version`, instead of only per-scope - Publish containers and scheduled task scopes as docker images - Remove unused cloudwatch annotations from deployment objects From ab6436082a8bcbb76caad7e1e2c72e968f51d7a2 Mon Sep 17 00:00:00 2001 From: Pablo Vilas Date: Wed, 2 Sep 2026 15:37:44 -0300 Subject: [PATCH 2/3] docs(changelog): tracing for k8s scope and deployment workflows (#215) --- CHANGELOG.md | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index eee43447..9c0a0aca 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -6,11 +6,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [Unreleased] -- k8s scopes now report their lifecycle and deployment operations to the nullplatform tracing API, so the dashboard shows the same step-by-step timeline for these scopes as for native ones: create, update and delete of the scope, and every deployment action (initial, blue/green, switch traffic, finalize, rollback, delete, diagnose, kill instance, restart pods, pause/resume autoscaling, set desired instance count). Each step shows its duration, long waits report live what they are waiting on (load balancer, DNS, instance health), and each workflow and step carries a human-readable title. Requires `NP_API_KEY` on the agent; without it, or without the bundled tracing SDK, behaviour is unchanged -- k8s scope deployments that fail now name the cause on the step where it happened, with what to do about it: image pull errors with the registry's message, out-of-memory kills, crash exit codes with the application's last log lines, failing startup/readiness probes with the path and response detected. The same message reaches the workflow engine's failure report instead of a bare exit status -- k8s **diagnose** checks now state their finding on their own step (summary, severity, affected pods, recommendation), so the deployment page shows what a check found instead of a list of green steps -- k8s scope and deployment workflows declare a per-provider job identity (`entity`, `operation`, `scope.provider`), so the dashboard can show the pending checklist for an operation before the agent picks the command up -- k8s scope deployments record lineage on the trace: the namespace, deployment, service and ingress they applied or removed, the DNS record and load balancer they depend on, and the build asset they deployed +- k8s scopes now show live, step-by-step progress in the dashboard when a scope is created, updated or deleted, and for every deployment action (deploy, switch traffic, finalize, rollback, delete, diagnose, kill instance, restart pods, pause/resume autoscaling, set instance count): each step with its duration, and long waits (load balancer, DNS, instance health) saying what they are waiting for. Requires `NP_API_KEY` on the agent; without it everything works as before +- Failed k8s deployments now explain why on the step that failed, and what to do about it: image pull errors with the registry's message, out-of-memory kills, crash exit codes with the application's last log lines, and failing health checks with the path and response detected +- k8s **diagnose** now shows what each check found (summary, severity, affected pods, recommended action) instead of a list of green steps - k8s scopes can now pin the traffic-manager sidecar version cluster-wide via the container-orchestration provider's `traffic_manager.version`, instead of only per-scope - Publish containers and scheduled task scopes as docker images - Remove unused cloudwatch annotations from deployment objects From bd4e2a60263d68357cd7c88de932908c98b89cd2 Mon Sep 17 00:00:00 2001 From: Federico Maleh Date: Wed, 2 Sep 2026 15:55:39 -0300 Subject: [PATCH 3/3] update changelog --- CHANGELOG.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index 9c0a0aca..85d10fb7 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -5,7 +5,7 @@ All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). -## [Unreleased] +## [1.16.0] - 2026-09-02 - k8s scopes now show live, step-by-step progress in the dashboard when a scope is created, updated or deleted, and for every deployment action (deploy, switch traffic, finalize, rollback, delete, diagnose, kill instance, restart pods, pause/resume autoscaling, set instance count): each step with its duration, and long waits (load balancer, DNS, instance health) saying what they are waiting for. Requires `NP_API_KEY` on the agent; without it everything works as before - Failed k8s deployments now explain why on the step that failed, and what to do about it: image pull errors with the registry's message, out-of-memory kills, crash exit codes with the application's last log lines, and failing health checks with the path and response detected - k8s **diagnose** now shows what each check found (summary, severity, affected pods, recommended action) instead of a list of green steps