## Description
When multiple organizations have templates with the same name, the
Prometheus `/metrics` endpoint returns HTTP 500 because Prometheus
rejects duplicate label combinations. The three `coderd_insights_*`
metrics (`coderd_insights_templates_active_users`,
`coderd_insights_applications_usage_seconds`,
`coderd_insights_parameters`) used only `template_name` as a
distinguishing label, so two templates named e.g. `"openstack-v1"` in
different orgs would produce duplicate metric series.
This adds `organization_name` as a label to all three insight metric
descriptors to disambiguate templates across organizations.
## Changes
**`coderd/prometheusmetrics/insights/metricscollector.go`**:
- Added `organization_name` label to all three metric descriptors
- Added `organizationNames` field (template ID → org name) to the
`insightsData` struct
- In `doTick`: after fetching templates, collect unique org IDs, fetch
organizations via `GetOrganizations`, and build a
template-ID-to-org-name mapping
- In `Collect()`: pass the organization name as an additional label
value in every `MustNewConstMetric` call
**`coderd/prometheusmetrics/insights/testdata/insights-metrics.json`**:
Updated golden file to include `organization_name=coder` in all metric
label keys.
Fixes#21748
Fixes#22030
## Problem
When a template has `require_active_version = true` and a workspace is
outdated, the web UI always shows "Update and start" as the **only**
button (for all users including admins), but `coder start` starts with
the old version. For admins, this silently succeeds on the stale
version. For non-admins, it goes through a clunky 403→retry path. This
also affects the VS Code extension, which calls `coder start --yes`
under the hood.
## Root Cause
`buildWorkspaceStartRequest()` in `cli/start.go` checks
`workspace.AutomaticUpdates == "always"` but ignores
`workspace.TemplateRequireActiveVersion`. The server-side autostart
already ORs both settings together:
```go
// coderd/autobuild/lifecycle_executor.go
func useActiveVersion(opts, ws) bool {
return opts.RequireActiveVersion || ws.AutomaticUpdates == "always"
}
```
The CLI was missing the `RequireActiveVersion` check.
## Fix
Add `workspace.TemplateRequireActiveVersion` to the existing OR
condition:
```go
// Before:
if workspace.AutomaticUpdates == codersdk.AutomaticUpdatesAlways || action == WorkspaceUpdate {
// After:
if workspace.AutomaticUpdates == codersdk.AutomaticUpdatesAlways || workspace.TemplateRequireActiveVersion || action == WorkspaceUpdate {
```
Now `coder start` and `coder restart` proactively use the active
template version when `require_active_version` is set, matching the web
UI and server autostart behavior. The 403→retry fallback remains as a
safety net but is no longer the primary path for any user.
## Testing
Updated `enterprise/cli/start_test.go` — all user types (owner, template
admin, ACL admin, group ACL admin, member) now expect the active version
when `require_active_version` is set, and verify the 403→retry message
does NOT appear.
## Summary
Custom roles that can create workspaces on behalf of other users need to
be able to list users to populate the owner dropdown in the workspace
creation UI. Previously, this required a separate `user:read`
permission, causing the dropdown to fail for custom roles.
## Changes
- Modified `GetUsers` in `dbauthz` to check if the user can create
workspaces for any owner (`workspace:create` with `owner_id: *`)
- If the user has this permission, they can list all users without
needing explicit `user:read` permission
- Added tests to verify the new behavior
## Testing
- Updated mock tests to assert the new authorization check
- Added integration tests for both positive and negative cases
Fixes#18203
## Summary
Fixes flaky `TestServer/BuiltinPostgres` test caused by port conflicts
in CI.
## Fix
Increase retry attempts from 3 to 10 for better odds when port conflicts
occur.
Fixes https://github.com/coder/internal/issues/1017
Fixes https://github.com/coder/coder/issues/18350
I attempted the route of relying on just the session env vars, in hopes
that this issue was fixed in Toolbox and the process name matching was
no longer need, but it was not a fruitful endeavor and it seems to be
using the same connection logic as it did in gateway, just with new
binary and flag names.
## Summary
- Updated documentation references to use the correct registry path
structure
- Changed URLs from incorrect paths to the correct format:
- Modules main page: registry.coder.com/modules
- Specific modules: registry.coder.com/modules/coder/{module-name}
## Test plan
- Verify that all documentation links to modules point to the correct
URL structure
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Edward Angert <EdwardAngert@users.noreply.github.com>
Fixes an issue where old ssh configs that use the
`owner--workspace--agent` format will fail to properly use the `coder
ssh` command since we migrated off the `coder vscodessh` command.
Part of https://github.com/coder/internal/issues/150
- Stop creating the VPC since we rely on an existing VPC in the
scaletest GCP project now
- Add prometheus remote writer to the new prometheus deployment
Closes https://github.com/coder/internal/issues/148
This has been validated to have working proxies with the `small`
scenario.
- [x] multi-region gcp infrastructure
- [x] use cloudflare provider to automate dns entries
- [x] automate proxy registration
- [x] multi-regional proxies
- [x] move scenarios into locals
This excludes the infrastructure for `cert-manager` and `otel-collector`
and those will be included in followup PRs.
Part of https://github.com/coder/coder/issues/15176
I originally kept this the same because I wanted to be conservative
about when we start dropping activity, but this is proving to be a
problem when using `coder ssh` with `--usage-app=disabled`. Because the
workspace agent still counts this as a connection (I think it still
should so it's counted somewhere) but not as a SSH / IDE session. This
leads to background ssh tasks that want to be untracked still continuing
to bump activity when it shouldn't. This makes it so we have to have an
explicit session to bump activity.
What this changes:
- Unhides the `--key` flag on provisioner start
- Deprecates and hides `provisionerd` command group in favor of
`provisioner(s)`
- Removes org id from `coder provisioner keys list`
Garrett Delfosse