From a201af9d29a5f4c0e313d5788d87969596276530 Mon Sep 17 00:00:00 2001 From: jwright Date: Mon, 17 Aug 2026 15:50:30 +1000 Subject: [PATCH] supply chain: tiered mirrors (central cloud mirror -> site mirrors) in prose, ledger card and blueprint --- assets/dist/delivery-flow.svg | 39 +++++++++++++++++------------- content/course-VII-supply-chain.md | 5 ++++ pilot/arc.tpl.html | 10 +++++++- 3 files changed, 36 insertions(+), 18 deletions(-) diff --git a/assets/dist/delivery-flow.svg b/assets/dist/delivery-flow.svg index 3cf459e..26906ff 100644 --- a/assets/dist/delivery-flow.svg +++ b/assets/dist/delivery-flow.svg @@ -1,4 +1,4 @@ - + @@ -67,26 +67,31 @@ ARTEFACT - - CI builds - once, only once - - ACR upstream - tag@sha256 minted - - zot mirror - the fleet pulls once - - node pulls - by digest + + CI builds + once, only once + + ACR upstream + tag@sha256 minted + + central mirror + one cloud cache, fronts upstream + + site mirror + zot at the edge + + node pulls + by digest - - - - + + + + + promotion is a copy of the same digest - if the digest changed, it is a new candidate, not a promotion + the tiers layer the load: upstream sees one consumer, each site sees one hop, nodes never leave the LAN SECRET diff --git a/content/course-VII-supply-chain.md b/content/course-VII-supply-chain.md index 4d7fd4c..57621a8 100644 --- a/content/course-VII-supply-chain.md +++ b/content/course-VII-supply-chain.md @@ -42,6 +42,11 @@ from Move 1 is what makes the mirror actually catch the traffic. FIELD NOTE: a mirror can also serve a STALE tag->digest mapping while upstream moved the tag. Pin by digest and the whole chain becomes content-addressed - the mirror cannot be wrong about a hash. +TIERING (added 2026-08-17, owner ask): at real fleet scale the mirror itself tiers - a +central pull-through in the cloud fronts upstream once; per-site mirrors pull from the +centre; nodes pull from their site over the LAN. The load fans out in layers instead of +every site hammering upstream at once. Say it: mirrors tier - the load fans out in +layers, never all at once. ## Move 3 - Build once, promote by copy The anti-pattern: rebuild the image for each environment ("the prod build"). Every diff --git a/pilot/arc.tpl.html b/pilot/arc.tpl.html index 5cdb2da..70389c4 100644 --- a/pilot/arc.tpl.html +++ b/pilot/arc.tpl.html @@ -237,6 +237,13 @@ image mirror rules, and carries a sharp edge: digest-mirror rules rewrite digest pulls only, so a by-tag pull silently skips the mirror. The pinning habit is what makes the mirror actually catch the traffic.

+

At real fleet scale the mirror itself tiers: a central mirror in the cloud fronts + upstream once, and every site's mirror pulls from the centre rather than from upstream + directly. A new image ripples outward in layers - upstream to the centre, centre to each + site as it asks, site to its nodes over the LAN - instead of every site hammering upstream + at the same moment. Upstream sees one consumer; each site sees one hop; the nodes never + leave the building.

+

Mirrors tier: the load fans out in layers, never all at once.

Build once, promote by copy

Every rebuild is a different artefact - different digest, untested by the stages before it. So build once, then promote the same digest through environments by copying, registry @@ -353,7 +360,8 @@

zot

Where the fleet pulls from - a pull-through mirror between the cluster and the - internet, and the control point for what enters.

+ internet, and the control point for what enters. At fleet scale it tiers: one central + mirror in the cloud fans out to per-site mirrors, layering the load.

Leans on it: every image pull on every node - boot, reschedule, scale-up, recovery.

When it is down: the sharpest edge. Upstream down + mirror up = nobody