## Problem
Site-wide admins (e.g., Owners) could not use `coder create --org <org>`
to create workspaces in organizations they are not members of. The error
was:
```
$ coder create my-workspace -t docker --org data-science
error: organization "data-science" not found, are you sure you are a member of this organization?
```
This was inconsistent with the web UI, where Owners can create
workspaces in any organization.
## Root Cause
The CLI's `OrganizationContext.Selected()` function only checked the
user's membership list, ignoring site-wide RBAC permissions that grant
Owners access to all organizations.
## Solution
Added a fallback in `OrganizationContext.Selected()` that fetches the
org directly via the API when not found in the membership list. This
works because the API endpoint applies RBAC filtering, allowing Owners
to read any org.
## Impact
This fixes `coder create --org` and all other CLI commands that use
`OrganizationContext.Selected()` (29+ commands), including:
- `coder templates push --org <any-org>`
- `coder organizations members add --org <any-org>`
- `coder provisioner list --org <any-org>`
## Testing
Added `TestEnterpriseCreate/OwnerCanCreateInNonMemberOrg` which:
- Creates an Owner user who is NOT a member of a second org
- Verifies they can create a workspace there using `--org`
- Properly fails without the code fix, passes with it
---
*This PR was generated by [mux](https://mux.coder.com) but reviewed by a
human.*
This PR adds some metrics to help identify job enqueue rates and
latencies. This work was initiated as a way to help reduce the cost of
the observation/measurement itself for autostart scaletests, which
impacts our ability to identify/reason about the load caused by
autostart. See: https://github.com/coder/internal/issues/1209
I've extended the metrics here to account for regular user initiated
builds, prebuilds, autostarts, etc. IMO there is still the question here
of whether we want to include or need the `transition` label, which is
only present on workspace builds. Including it does lead to an increase
in cardinality, and in the case of the histogram (when not using native
histograms) that's at least a few extra series for every bucket. We
could remove the transition label there but keep it on the counter.
Additionally, the histogram is currently observing latencies for other
jobs, such as template builds/version imports, those do not have a
transition type associated with them.
Tested briefly in a workspace, can see metric values like the following:
-
`coderd_workspace_builds_enqueued_total{build_reason="autostart",provisioner_type="terraform",status="success",transition="start"}
1`
-
`coderd_provisioner_job_queue_wait_seconds_bucket{build_reason="autostart",job_type="workspace_build",provisioner_type="terraform",transition="start",le="0.025"}
1`
---------
Signed-off-by: Callum Styan <callumstyan@gmail.com>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
## Description
This PR addresses database connection pool exhaustion during prebuilds
reconciliation by introducing two changes:
* `CanSkipReconciliation`: Filters out presets that don't need
reconciliation before spawning goroutines. This ensures we only create
goroutines for presets that will (_most likely_) perform database
operations, avoiding unnecessary connection pool usage.
* Dynamic `eg.SetLimit`: Limits concurrent goroutines based on the
configured database connection pool size (`CODER_PG_CONN_MAX_OPEN / 2`).
This replaces the previous hardcoded limit of 5, ensuring the
reconciliation loop scales appropriately with the configured pool size
while leaving capacity for other database operations.
## Changes
* Add `CanSkipReconciliation()` method to `PresetSnapshot` that returns
true for inactive presets with no running workspaces, no pending jobs,
or expired prebuilds.
* Add `maxDBConnections` parameter to `NewStoreReconciler` and compute
`reconciliationConcurrency` as half the pool size (minimum 1).
* Add `ReconciliationConcurrency()` getter method to `StoreReconciler`.
* Add `eg.SetLimit(c.reconciliationConcurrency)` to bound concurrent
reconciliation goroutines.
* Add `PresetsTotal` and `PresetsReconciled` to `ReconcileStats` for
observability.
* Add `TestCanSkipReconciliation` unit tests.
* Add `TestReconciliationConcurrency` unit tests.
* Add benchmark tests for reconciliation performance.
## Benchmarks
* `BenchmarkReconcileAll_NoOps`: Tests presets with no reconciliation
actions. All presets are filtered by `CanSkipReconciliation`, resulting
in no goroutines spawned and no database connections used.
* `BenchmarkReconcileAll_ConnectionContention`: Tests presets where all
require reconciliation actions. All presets spawn goroutines, but
concurrency is limited by `eg.SetLimit(reconciliationConcurrency)`.
* `BenchmarkReconcileAll_Mix`: Simulates a realistic scenario with a
large subset of inactive presets (filtered by `CanSkipReconciliation`)
and a smaller subset requiring reconciliation (limited by
`eg.SetLimit`).
Closes: https://github.com/coder/coder/issues/20606
Fixes all our Go file imports to match the preferred spec that we've _mostly_ been using. For example:
```
import (
"context"
"time"
"github.com/prometheus/client_golang/prometheus"
"golang.org/x/xerrors"
"gopkg.in/natefinch/lumberjack.v2"
"cdr.dev/slog/v3"
"github.com/coder/coder/v2/codersdk/agentsdk"
"github.com/coder/serpent"
)
```
3 groups: standard library, 3rd partly libs, Coder libs.
This PR makes the change across the codebase. The PR in the stack above modifies our formatting to maintain this state of affairs, and is a separate PR so it's possible to review that one in detail.
The implementation for prebuilt workspaces is complex and conversations
regarding edge cases and bugs frequently get bogged down by minutiae,
because it's hard to reason about the behaviour of the system.
To alleviate this, I've introduced otel tracing to the StoreReconciler
(see attached). We can now directly observe the behaviour of the
prebuilds system under load in order to inform our decisions.
Traces are terminated at the boundary between prebuilds and workspace
builder, because of prebuilt workspaces' "fire and forget" philosophy
and to prevent span explosion.
<img width="3024" height="1718" alt="image"
src="https://github.com/user-attachments/assets/f9b207be-8f2c-475e-98a8-46ef70bda446"
/>
Provisioner steps broken into smaller granular actions.
Changes:
- `ExtractArchive` moved to `init` request (was in `configure`)
- Writing `tfstate` moved to `plan` (was in `configure`)
- Moved most plan/apply outputs to `GraphComplete`
## Description
This PR introduces a `--preset` flag for the `create` command to allow
users to apply a predefined preset to their workspace build.
## Changes
- The `--preset` flag on the `create` command integrates with the
parameter resolution logic and takes precedence over other sources
(e.g., CLI/env vars, last build, etc.).
- Added internal logic to ensure that preset parameters override
parameters values during resolution.
- Updated tests and added new ones to cover these flows.
## Implementation logic
* If a template has presets and includes a default, the CLI will
automatically use the default when `--preset` is not specified.
* If a template has presets but no default, the CLI will prompt the user
to select one when `--preset` is not specified.
* If a template does not have presets, the CLI will not prompt the user
for a preset.
* If the user specifies a preset using the `--preset` flag, that preset
will be used.
* If the user passes `--preset None`, no preset will be applied.
This logic aligns with the behavior in the UI for consistency.
```
> coder create --help
USAGE:
coder create [flags] [workspace]
Create a workspace
- Create a workspace for another user (if you have permission):
$ coder create <username>/<workspace_name>
OPTIONS:
(...)
--preset string, $CODER_PRESET_NAME
Specify the name of a template version preset. Use 'none' to explicitly indicate that no preset should be used.
(...)
-y, --yes bool
Bypass prompts.
```
## Breaking change
**Note:** This is a breaking change to the create CLI command. If a
template includes presets and the user does not provide a `--preset`
flag, the CLI will now prompt the user to select one. This behavior may
break non-interactive scripts or automated workflows.
Relates to PR: https://github.com/coder/coder/pull/18910 - please
consider both PRs together as they’re part of the same workflow
Relates to issue: https://github.com/coder/coder/issues/16594
* chore: move multi-org endpoints into enterprise directory
All multi-organization features are gated behind "premium" licenses. Enterprise licenses can no longer
access organization CRUD.
Ethan