Your workspace is ready.
No datasets have been configured. Browse an accessible database to begin.
No datasets have been configured. Browse an accessible database to begin.
Follow scheduled collections from the first request to a durable raw response.
| Dataset | State | Last complete collection | Source freshness | Next due · UTC |
|---|
Follow collection performance and configured infrastructure resources, with separate sources and freshness.
CPU, memory and storage come from configured server-side infrastructure collectors. Pipeline provider, dataset and fixture/live filters do not apply here.
| Resource / source | State / setup | Last sample · UTC | CPU | Memory | Storage |
|---|---|---|---|---|---|
| No resource targets loaded. | |||||
Resource PagerDuty · status unavailable in this snapshot.
Each metric has its own observation time. Unavailable metrics stay unavailable; absence does not indicate zero usage.
| Metric | Latest measurement | Observed · UTC |
|---|
Configure targets, DigitalOcean monitoring credentials and any monitoring agent in the private server runtime. No token, database password or connection string is entered in this dashboard. Missing credentials, unsupported metrics and collector failures are reported explicitly. An ingest heartbeat does not establish host CPU, RAM or database storage usage.
DigitalOcean Standard PostgreSQL uses its external Prometheus endpoint; Advanced has different metric availability. Droplet RAM and disk measurements require its monitoring agent. See the infrastructure monitoring guide ↗.
Accepted records and captured response bytes measure collection output. Source freshness is separate from a successful collection. Select a pipeline for its conditions and latest run.
| Dataset / provider | Health / mode | Ingest rate | Window volume | Run success / latency | Collection / source age | Limits / errors |
|---|---|---|---|---|---|---|
| No snapshot loaded. | ||||||
| Time · UTC | Severity / type | Dataset / run | Event |
|---|---|---|---|
| No snapshot loaded. | |||
Bounded event history from stored request attempts, run outcomes and operational signals. Search applies to the recent events returned for this window.
Integration readiness is read from the server. Routing keys remain in the server's secret configuration.
Configure the PagerDuty Events API v2 integration in the private server runtime file using a routing key stored in your secrets manager. Enable the integration and run the independent outbox process so it can evaluate pipeline conditions and deliver notifications. The default alert mode is live. This dashboard displays readiness and delivery status.
See monitoring setup and alert rules ↗ for exact configuration and fixture notification behavior.
Connect a REST API or WebSocket stream, or configure a supported provider profile. Save paused or explicitly start a verified connection.
Connect a REST JSON endpoint or a continuous WebSocket stream, preview its fields, and choose exactly what to store.
Queued, running and committed are separate states. Manual runs are new observations.
| Run | Dataset | Result | Scheduled · UTC | Pages | Action |
|---|
Read-only response metadata and bounded body previews. Partial and invalid responses remain inspectable.
| Response | Provider | Received · UTC | Validation | Bytes | Archive |
|---|
Inspect PostgreSQL databases, schemas and tables, and manage approved research databases.
| Database | Owner | Size | Connections | Access | Inspection |
|---|---|---|---|---|---|
| Open Databases to load the inventory. | |||||
Checking the capabilities of this database and your account…
| Object | Kind | Owner / table |
|---|
Inspect its catalog metadata, then use the existing table browser for records and exports.
Structured operations apply only to owned, registered research databases. Protected application/system databases remain read-only. Review the exact target before applying changes.
Activity is a bounded snapshot of sessions owned by the database management role. Query text is redacted. The current inspection session cannot be canceled; other queries require explicit server verification.
| PID / started | Role / connection | State / wait | Query timing / redacted text | Action |
|---|
| PID | Object / type | Mode | Granted / blocking |
|---|
Run bounded ANALYZE, VACUUM or REINDEX operations where explicitly permitted. These controls do not grant cluster or operating-system administration.
| Setting | Value | Context / source |
|---|
Scripts are saved explicitly on the server for your account. SQL and result data are never saved to browser storage. Loading a script does not execute it.
| Script | Updated · UTC | Manage |
|---|
History contains command hashes and outcomes, without SQL literals or result bodies.
| Time · UTC | Action / hash | Outcome | Details |
|---|
Export the selected database through the existing backup action. Restore a trusted PostgreSQL custom-format archive into a new registered research database; existing databases are never overwritten.
| Destination | State | Archive / created | Result | Action |
|---|
Visibility and grant management depend on the configured database role and superadmin access. Application users and roles are managed separately in Users & roles.
| Role | Capabilities | Membership |
|---|
| Schema / table | Role | Privileges |
|---|
Checking whether the optional private role-management connection is configured.
| Group | Registry state | Members | Manage |
|---|
| Schema | Tables |
|---|
| Relation | Type | Estimated rows | Total size |
|---|
Inventory row counts are PostgreSQL estimates. Only tables exposed by your database access and role are shown. Select a relation for an exact filtered count and its stored rows.
| Column | PostgreSQL type | Nullable | Key / generated | Default |
|---|
Export all rows matching the applied exact-value and date criteria, across every page. This export excludes rows outside those criteria.
Deletion permanently removes the entire database, including every schema, table and stored record. A typed confirmation is required.
Publish selected database columns through a read-only API with scoped keys.
| Publication | Source | Approved columns | State | TTL · seconds | Controls |
|---|---|---|---|---|---|
| Open API access to load publications. | |||||
Select a publication to see a request with a placeholder key.
Use HTTPS through the agreed access route. Numeric values and exact JSON are returned as strings. Limits, sorting and exact filters apply only to published columns.
| Key | Scope | Limits | Expires · UTC | Usage | State | Control |
|---|
Lifetime totals are retained. Daily history covers 90 days; recent request details are bounded to 30 days and 1,000 per key.
| Day | Requests | Cache hits | Errors | Bytes |
|---|
| Time · UTC | Key / endpoint | Status | Cache | Bytes | Duration |
|---|
Profiles, schedules, stable IDs and credential references in versioned JSON. Provider credentials and market data are excluded.
Validate the full document and inspect changes first. Imported schedules start paused. Re-importing the same definitions keeps their IDs.
Only superadmins manage users and roles. Changes take effect on subsequent requests.
| Username | Status | Assigned roles | Updated · UTC | Manage |
|---|
Issue a single-use code with reviewed roles. Codes are displayed once for you to share privately; they are not emailed or added to links. Superadmin access requires a separate manual promotion.
Select and save this code through your approved private channel. It cannot be retrieved after dismissal, refresh or navigation.
| Label / invite | Status | Assigned roles | Expiry / creator | Used by | Manage |
|---|---|---|---|---|---|
| Open Users & roles to load invites. | |||||
Latest invitations; codes are never included in this list.
Read only, Admin and Superadmin are predefined. Custom roles grant individual actions for exact database, schema and table names, or * for every resource at that level. Grants are additive.