From 98bcd5cc4941cc6475a6e4a9823442a53995b6ef Mon Sep 17 00:00:00 2001 From: jochen Date: Mon, 24 Aug 2026 00:07:22 +0200 Subject: [PATCH] =?UTF-8?q?Measure=20the=20lab's=20inner=20loop=20?= =?UTF-8?q?=E2=80=94=20it=20is=20too=20slow,=20for=20a=20fixable=20reason?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The lifecycle design closed on a question that was measurable rather than arguable, so it was measured. One virtual machine on a workstation with hardware virtualisation and NVMe. Raising: the launch call returns in 3.4s, the machine is actually usable after 14.3s. The gap is a design constraint — raise must wait for the second number, because reporting the first would be transport reported as effect, which is the mesh's own recurring failure. Snapshot: 9.9s and 1.6 GB for a 1.5 GB instance. A dir snapshot is a full copy; nothing is shared. Restore: 10.4s, usable again after 20.1s. The second snapshot exceeded two minutes and never completed. That is the more troubling number: snapshot cost here is not merely high, it is unpredictable, and a loop with a variable multi-minute step is one nobody trusts. Projected to a four-machine scenario, a reset-and-rerun cycle is about a minute and a half at best and unbounded at worst, before any of the mesh's own work begins. That is too slow for an inner loop, and ADR 0029's whole argument — that making the bootstrap path the inner loop turns the least-exercised code into the most-exercised — holds only while resetting is cheap. The cause is not virtual machines. Hardware virtualisation is present and machines boot in fourteen seconds. It is that the daemon offers exactly one storage driver, dir, which has no copy-on-write and therefore no cheap snapshot. The btrfs kernel module is available; btrfs-progs is simply not installed, which is the entire reason the driver is absent. The copy-on-write comparison was deliberately NOT run, because running it would mean installing a package by hand — which the rules forbid and which would have made the measurement unreproducible. So the honest statement is that the current configuration is too slow and the likely fix is known but unverified, rather than that btrfs fixes it. --- .../010-lab-inner-loop-cost/00-overview.md | 55 +++++++++++ .../010-lab-inner-loop-cost/measurements.md | 98 +++++++++++++++++++ 03-DESIGN/01-to-be/03-scenario-lifecycle.md | 12 ++- 3 files changed, 162 insertions(+), 3 deletions(-) create mode 100644 01-RESEARCH/010-lab-inner-loop-cost/00-overview.md create mode 100644 01-RESEARCH/010-lab-inner-loop-cost/measurements.md diff --git a/01-RESEARCH/010-lab-inner-loop-cost/00-overview.md b/01-RESEARCH/010-lab-inner-loop-cost/00-overview.md new file mode 100644 index 0000000..bd52bde --- /dev/null +++ b/01-RESEARCH/010-lab-inner-loop-cost/00-overview.md @@ -0,0 +1,55 @@ +--- +status: active +initiated: 2026-08-24 +touches: + - 03-DESIGN/01-to-be/03-scenario-lifecycle.md + - 02-DECISIONS/0016-a-lab-node-is-a-virtual-machine.md + - 02-DECISIONS/0029-the-labs-first-scenario-has-no-pipeline.md +became: [] +--- + +# 010 — What the lab's inner loop actually costs + +## What is being investigated + +[The lifecycle design](../../03-DESIGN/01-to-be/03-scenario-lifecycle.md) closes on an open +question that is measurable rather than arguable: + +> **What a snapshot costs.** Whole-scenario snapshots of several machines are the operation the +> inner loop repeats most, so their cost sets the loop's speed. If restoring is slow, the loop +> is slow, and everything above is theory. + +Measured on a workstation, 2026-08-24. Numbers in [`measurements.md`](measurements.md). + +## Why it matters + +[ADR 0029](../../02-DECISIONS/0029-the-labs-first-scenario-has-no-pipeline.md) makes the +bootstrap scenario the inner development loop for tiers 0 and 1 — the argument being that +raising a node from nothing stops being the least-exercised path and becomes the most-exercised +one. **That argument is only true if raising and resetting are cheap.** A loop that costs +minutes is a loop people avoid, and the least-exercised path stays least-exercised. + +## Status + +Measured, and the finding is a blocker rather than a data point. + +**A snapshot on the current host is a full copy of the machine's disk.** 1.6 GB and ten seconds +for one small virtual machine at best — and over two minutes when observed a second time. Cost +scales with the number of machines and the size of their disks, not with what changed. + +The cause is not virtual machines and not incus. It is that the host offers incus exactly one +storage driver, `dir`, which has no copy-on-write and therefore no cheap snapshot. The kernel +supports btrfs; the userspace tool that would let incus use it is simply not installed. + +So the question *"is the lab's inner loop fast enough"* currently answers itself the wrong way, +for a reason that is one declared package away from being fixed — and declaring packages is +something the mesh already does. + +## Open questions + +| Question | Why it matters | +|---|---| +| How much does a copy-on-write pool actually improve it? Expected to be near-instant snapshots and delta-sized storage, but **expected is not measured**. | The whole inner-loop argument rests on the answer. | +| Why was the second snapshot more than twelve times slower than the first? | If snapshot cost is unpredictable rather than merely high, that is worse — a loop with a variable multi-minute step is one nobody trusts. | +| Does a scenario snapshot need the machines stopped? | Stateless snapshots of a running virtual machine capture the disk but not memory. Whether a mesh restored that way is coherent is not established. | +| What is the cost at scenario scale — four machines rather than one? | Only single-machine numbers were taken. If the operation is serial, four machines is four times the wait. | diff --git a/01-RESEARCH/010-lab-inner-loop-cost/measurements.md b/01-RESEARCH/010-lab-inner-loop-cost/measurements.md new file mode 100644 index 0000000..9189c07 --- /dev/null +++ b/01-RESEARCH/010-lab-inner-loop-cost/measurements.md @@ -0,0 +1,98 @@ +--- +effort: 010-lab-inner-loop-cost +updated: 2026-08-24 +--- + +# Measurements + +Taken 2026-08-24 on a workstation with hardware virtualisation available, an NVMe-backed ext4 +root, and 300 GB free. One virtual machine, 1 GiB memory, 2 CPUs, from a cached distribution +image. + +## The environment, before anything ran + +| Fact | Value | Consequence | +|---|---|---| +| Hardware virtualisation | present | virtual machines run at native speed; the choice in [ADR 0016](../../02-DECISIONS/0016-a-lab-node-is-a-virtual-machine.md) is not paying an emulation penalty | +| Storage drivers the daemon offers | **`dir` only** | no copy-on-write, therefore no cheap snapshot | +| Host filesystems | ext4 throughout | nothing copy-on-write to put a pool on | +| btrfs kernel module | **available** | the kernel can do it | +| `btrfs-progs` | **not installed** | which is the entire reason the driver is absent | + +The last two rows are the finding. The daemon advertises only `dir` because the userspace tool +for anything better is missing — not because the host cannot do better. + +## Raising a machine + +| Step | Time | +|---|---| +| launch call returns | **3.4 s** | +| machine actually usable — a command executes on it | **14.3 s** | + +The gap matters for the lifecycle design: `raise` returning is not the same as the scenario +being ready, so the verb has to wait for the second number, not report the first. Reporting +the first would be the mesh's own recurring failure — transport reported as effect. + +## Snapshot and restore + +| Operation | Time | Disk | +|---|---|---| +| snapshot, first | **9.9 s** | **+1.6 GB** | +| snapshot, second | **> 120 s — did not complete** | — | +| restore call returns | **10.4 s** | — | +| machine usable again | **20.1 s** total | — | + +Instance on disk before snapshotting: 1.5 GB. Snapshot directory afterwards: 1.6 GB. **A `dir` +snapshot is a full copy** — the storage cost equals the instance, and nothing is shared. + +Implied copy throughput on the first snapshot is roughly 160 MB/s, which is far below what the +underlying NVMe can do and is consistent with a real, durable copy rather than a metadata +operation. + +**The second snapshot is the more troubling number.** It exceeded two minutes and was still +running when the observation was cut off; only the first snapshot exists. Whatever the cause — +page cache exhausted by the preceding restore, writeback contention — the practical +consequence is that snapshot cost here is **not merely high, it is unpredictable**. + +## What this projects to + +A four-machine scenario, taking the optimistic single-machine numbers and assuming the +operations are serial: + +| | one machine | four machines | +|---|---|---| +| raise, to usable | 14 s | ~57 s | +| snapshot | 10 s, 1.6 GB | ~40 s, 6.4 GB | +| restore, to usable | 20 s | ~80 s | + +A reset-and-rerun cycle is therefore **around a minute and a half at best**, and unbounded at +worst, before any of the mesh's own work begins. + +## The judgement + +**This is too slow for an inner loop**, and the reason is not the design. + +[ADR 0029](../../02-DECISIONS/0029-the-labs-first-scenario-has-no-pipeline.md) argues that +making the bootstrap path the inner development loop turns the least-exercised code in the +system into the most-exercised. That argument holds only while resetting is cheap. At a minute +and a half a cycle, with occasional multi-minute stalls, the loop is one a person works around +— and the path stays under-exercised for exactly the reason it always was. + +Nothing about virtual machines causes this. Hardware virtualisation is present and the machines +boot in fourteen seconds. **The cost is entirely the storage driver**, and the driver is absent +because one userspace package is not installed on the host. + +The mesh already has the mechanism for that: a module declares a package, and a hook makes it a +working capability — which is precisely what was just done for the virtualisation daemon +itself, and what [`04-ISSUES/007`](../../04-ISSUES/007-an-installed-package-is-not-a-capability/00-report.md) +is about. + +## Not measured, and it matters + +The copy-on-write comparison **was not run**, because running it would mean installing a package +by hand, which the mesh's rules forbid and which would have made the measurement unreproducible +anyway. Copy-on-write snapshots are expected to be near-instant with storage proportional to +what changed. That expectation is well founded and is still an expectation. + +Until it is measured, the correct statement is: *the current configuration is too slow, and the +likely fix is known but unverified.* diff --git a/03-DESIGN/01-to-be/03-scenario-lifecycle.md b/03-DESIGN/01-to-be/03-scenario-lifecycle.md index 3ca2623..8a06df0 100644 --- a/03-DESIGN/01-to-be/03-scenario-lifecycle.md +++ b/03-DESIGN/01-to-be/03-scenario-lifecycle.md @@ -48,6 +48,10 @@ The order is not arbitrary — each step needs the one before it to exist: 5. **Placement.** Artifacts onto machines. 6. **Snapshot**, if the declaration named one. +`raise` must wait for the machines to be **usable**, not for the call to return. Measured, those +are 14 seconds apart for a single machine — and reporting the earlier number would be the mesh's +own recurring failure, transport reported as effect. + Raising is **convergent, not incremental**: raising an instance that already exists brings it to the declared state rather than failing or duplicating. That is the same model the mesh itself uses, and a lab that behaved differently from the thing it tests would be teaching the wrong @@ -122,9 +126,11 @@ decision rather than a second implementation. ## Open -- **What a snapshot costs.** Whole-scenario snapshots of several machines are the operation the - inner loop repeats most, so their cost sets the loop's speed. If restoring is slow, the loop - is slow, and everything above is theory. +- **What a snapshot costs — measured, and the answer is a blocker.** On the current host a + snapshot is a full copy: 1.6 GB and ten seconds for one small machine, over two minutes when + observed again, projecting to roughly a minute and a half for a four-machine reset-and-rerun + cycle. The cause is the storage driver, not virtual machines. See + [research 010](../../01-RESEARCH/010-lab-inner-loop-cost/measurements.md). - **Instance naming.** A declaration is a kind and instances are many; how they are named decides whether a person can find the one they left standing yesterday. - **What survives `destroy`.** Logs and captures are the output of a failed run, so destroying