Skip to content
Site Qwality Site Qwality Site Qwality Docs

Import monitors from a CSV

A CSV adds many monitors at once, one monitor per row. Before anything is created you see what each row will do: create a monitor, skip a duplicate, or stay out with the reason. A bad row never stops the rest of the file.

Moving from UptimeRobot, Pingdom, StatusCake or Better Stack? Import from that tool with its API key instead, with no file to build.

In the dashboard

Section titled “In the dashboard”

  1. On the dashboard, click Import next to Add monitor, open the CSV file tab, click Choose a CSV file and pick your file. The tab lists the columns and has an example file to download. A new account can also import from the first-monitor screen.

  2. Check the preview. Each row shows Will be created, Skipped with the reason, or Not imported with what is wrong. Nothing has been created yet.

  3. Click the Import button, which names how many rows are marked Will be created. Fix any rows with a problem and import the file again; the monitors you just created are skipped.

Example

Section titled “Example”
name,url,check_type,interval,tags
Homepage,https://example.com,,300,prod
API health,https://api.example.com/health,,60,"prod,api"
Postgres,db.example.com:5432,tcp,,
Edge router,ping://edge.example.com,,,
Nightly backup,,heartbeat,86400,ops
Main zone,example.com,dns,,

The first line names the columns, in any order. Column names ignore case, and blank lines are skipped.

Columns

Section titled “Columns”
ColumnAlso acceptedApplies toDescription
urluri, domainHTTP, TCP, ping, DNSWhat to check. TCP takes host:port or tcp://host:port, ping takes a host or ping://host, and DNS takes the zone (from a URL, only the host is kept). Leave it blank on heartbeat rows.
check_typetypeAllhttp, tcp, ping, heartbeat (or cron) or dns. When blank, a tcp:// or ping:// URL sets the type and any other URL is http.
namefriendly_nameAllThe monitor’s name. Heartbeat rows need one.
intervalinterval_seconds, run_interval_secondsAllSeconds between checks, in whole minutes up to 86400. For a heartbeat, how often the job pings. Default 300, or 3600 for DNS, which also needs at least 300.
methodHTTPThe request method. Default GET.
keywordkeyword_searchHTTPText the response must contain.
json_pathHTTPA JSON path to read from the response.
json_path_expectedexpectedHTTPThe value json_path must return.
tagsAll but DNSTags separated by , or ;. Quote the cell when it holds a comma.

Imported HTTP checks start with SSL/TLS, domain expiry and page speed checks off; turn them on per monitor. Every imported monitor alerts your default notification group. Once a heartbeat monitor exists, open it to copy its ping URL. Browser checks need a script, so create them in the dashboard or with the API.

Duplicates are skipped

Section titled “Duplicates are skipped”

A row is skipped when:

  • its URL matches one of your monitors. The scheme and host ignore case and a trailing slash is ignored, but the path must match exactly;
  • it is a heartbeat row named like one of your heartbeat monitors, ignoring case;
  • it is a DNS row for a zone you already monitor;
  • it repeats an earlier row of the same file.

So importing the same file again is safe: it creates only the rows that were not created before.

Limits

Section titled “Limits”
  • A file holds at most 500 monitor rows.
  • Each row is created just as if you added it by hand, so your plan’s monitor, heartbeat and DNS monitor limits and its shortest interval apply. The preview checks your plan’s monitor limit and shortest interval; the heartbeat and DNS monitor limits show up as a problem on the row when you import.
  • An import stops creating after about 20 seconds. Rows it did not reach show Not imported yet. Click Import the rest, or import the same file again, and the monitors already created are skipped.

With the API

Section titled “With the API”

With an API key exported as SITEQWALITY_API_KEY, send the file’s text to POST /http/jobs/import. dry_run: true reports what would happen and creates nothing:

Terminal window
curl -X POST https://api.siteqwality.com/http/jobs/import \
  -H "Authorization: Bearer $SITEQWALITY_API_KEY" \
  -H "Content-Type: application/json" \
  -d "$(jq -n --rawfile csv monitors.csv '{csv: $csv, dry_run: true}')"

Send it again without dry_run to create the monitors. The response has one entry per row:

{
  "data": {
    "dry_run": true,
    "created": 0,
    "would_create": 1,
    "skipped": 1,
    "failed": 1,
    "not_attempted": 0,
    "errors": ["line 4: interval is not a number"],
    "rows": [
      { "line": 2, "status": "would_create", "check_type": "http", "name": "Homepage", "target": "https://example.com", "id": null, "message": null, "source_id": null, "warnings": [] },
      { "line": 3, "status": "skipped", "check_type": "http", "name": "Docs", "target": "https://example.com/docs", "id": null, "message": "A monitor for this url already exists", "source_id": null, "warnings": [] },
      { "line": 4, "status": "failed", "check_type": null, "name": null, "target": null, "id": null, "message": "interval is not a number", "source_id": null, "warnings": [] }
    ]
  }
}

status is would_create (dry run only), created (with the new monitor’s id), skipped, failed or not_attempted. target is the URL, the heartbeat’s name or the DNS zone. errors repeats each failed row as line N: reason.

A file that cannot be read at all returns 400 with the reason and creates nothing: an empty file, no url or check_type column, no monitor rows, or more than 500 rows.

Privacy Cookie preferences