API

Most data collected and aggregated for your app can be accessed through our REST API. Use it to build custom dashboards, integrations, explore data in ways not possible through Monti APM's UI, or give access to the data in Monti APM to your agents. The API is available for all users and apps.

This REST API replaces our former GraphQL API, and provides more data with more flexible ways to query.

Stability

While we anticipate no large changes, the API is currently not considered stable as we continue to test and improve the API and the query system it uses. It is ready to be used to explore your app's data (personally or through an agent), but writing deeper integrations might require minor updates until we stabilize the API. Any breaking changes will be announced on our changelog.

Authenticate

You can get an API token in your account settings. Each token can be scoped to a single app or to access data for all of your apps.

The REST API uses bearer auth. Set the Authorization header to this, of course replacing <api token> with your API token.

Bearer <api token>

Routes

The base URL is https://api.montiapm.com/v1/, or https://api-us.montiapm.com/v1 for apps using the Galaxy integration.

List of metrics

A list of metrics and their filters, dimensions, and aggregations can be retrieved with a GET request to /metrics.

curl https://api.montiapm.com/v1/metrics \
  -H "Authorization: Bearer $MONTI_TOKEN"

For each metric, the first listed aggregation is the default aggregation. The default for specific metrics might change over time, so we recommend specifying the aggregation when querying metrics.

List of tag values

You can get a list of values for a tag with a POST request to /tag-values.

The body should be in the format of:

{
  "appIds": ["YOUR_APP_ID"],
  "tagId": "tag:host",
  "metricId": "methods.total"
}

A startTime and endTime can also be optionally provided, defaulting to the last hour.

The result looks like:

{
  "values": ["web-1", "web-2"]
}

The tag values can differ based on the metric, specifically with the collection the metric is from. For example, for the tag with the id tag:name will have different values for method and http metrics.

Provides a maximum of 1,000 values. appIds must have exactly one id; querying more than one app at a time is not supported at the moment.

Query

The metrics can be queried with a POST request to /query.

The body

There are 3 types of queries:

• timeseries provides a list of timestamps and values
• summary provides a single aggregated value for the whole time range (or per group when using dimensions)
• breakdown provides a list of groups for a metric ordered by the aggregated value of a metric

The response has an array for each field. The first index of each array is the data for the first result, the second index for the second result, etc.

{
  "timestamps": [1758600000000, 1758600060000, 1758600120000],
  "metrics": { "methods.total": [42.5, 38.1, 51.9] },
  "tags": { "tag:host": ["web-1", "web-1", "web-1"] }
}

• metrics has the value for each metric queried.
• tags has the value of each tag for each metric value/timestamp.
• timestamps is only provided for timeseries queries, and are in milliseconds.

The results will be aggregated into values for every minute, 30 minutes, or 3 hours, depending on the range between startTime and endTime in the query.

Timeseries have a limit of 200 values for each dimensions. If there are more values, the less common ones will be grouped into an "Other" dimension.

Example 1: how fast were the methods over the last hour?

curl https://api.montiapm.com/v1/query \
  -H "Authorization: Bearer $MONTI_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "appIds": ["YOUR_APP_ID"],
    "type": "timeseries",
    "metric": "methods.total"
  }'

Example 2: what are the 5 slowest methods?

{
  "appIds": ["YOUR_APP_ID"],
  "type": "breakdown",
  "metric": "methods.total",
  "dimensions": ["tag:name"],
  "limit": 5,
  "extraMetrics": ["methods.count"]
}

The response might look like:

{
  "metrics": {
    "methods.total": [812.4, 530.2, 301.7, 250.0, 199.3],
    "methods.count": [12, 340, 55, 8, 1020]
  },
  "tags": {
    "tag:name": ["generateReport", "search", "saveSettings", "exportCsv", "addTodo"]
  }
}

Reading down the columns: the method generateReport took 812 ms on average and was called an average of 12 times per minute and is the slowest. The second method, search, took 530ms and was called 340 times per minute, so optimizing it might have even more benefit.

Filtering

A filter is in the shape of:

{ "tagId": "tag:host", "operator": "is", "values": ["web-1", "web-2"] }

There are three operators available:

You can use more than one filter. Data has to be matched by all filters to be kept.

Not every filter works with every metric. Check GET /metrics to see which ones you can use.

List Traces

To get a list of traces, send a POST request to /traces with a body like:

{
  "type": "method",
  "appIds": ["YOUR_APP_ID"],
  "filters": [{ "tagId": "tag:name", "operator": "is", "values": ["generateReport"] }],
  "sortBy": "total",
  "order": "desc",
  "limit": 10
}

All of the options are:

There are 5 types of traces:

• method
• pub
• http
• job Also used for custom traces
• error

error traces can only be sorted by startTime. Other types can also be sorted by total, wait, db, compute, http, email, async, or fs.

The response returns a hasMore property, true if there are more pages of traces.

{
  "traces": [
    {
      "id": "eee2a942-24e3-4003-8c67-67ee5b810f7f",
      "name": "generateReport",
      "host": "web-1",
      "startTime": 1758598765000,
      "errored": false,
      "metrics": {"total": 2410, "db": 1900, "compute": 300, "wait": 210}
    }
  ],
  "hasMore": false
}

Single Trace

Send a GET request to /trace with the appId, type, and id query params:

https://api.montiapm.com/v1/trace?appId=YOUR_APP_ID&type=method&id=eee2a942-24e3-4003-8c67-67ee5b810f7f

The response will look like:

{
  "id": "abc123",
  "type": "method",
  "name": "generateReport",
  "host": "web-1",
  "startTime": 1758598765000,
  "errored": false,
  "metrics": { "total": 2410, "db": 1900, "compute": 300, "wait": 210 },
  "events": [
    ["start", 0],
    ["db", 1900, {"coll": "reports"}],
    ["compute", 300],
    ["wait", 210],
    ["complete", 0]
  ]
}

The events array contains an array for each event, in the format [type, duration, data, details]. Some details have a nested property with an array of child events.

Examples

10 most common errors in the last hour:

{
  "appIds": ["YOUR_APP_ID"],
  "type": "breakdown",
  "metric": "errors.count",
  "dimensions": ["tag:name"],
  "limit": 10,
  "extraMetrics": ["errors.lastSeen"]
}

List traces for a specific error

{
  "type": "error",
  "appIds": ["YOUR_APP_ID"],
  "filters": [
    { "tagId": "tag:name", "operator": "is", "values": ["Cannot read properties of undefined"]},
    { "tagId": "tag:type", "operator": "is", "values": ["method"]}
  ]
}

Get the 95th percentile Method response time:

{
  "appIds": ["YOUR_APP_ID"],
  "type": "summary",
  "metric": "methods.resTimeHistogram",
  "aggregation": "p95"
}

Response:

{
  "metrics": {"methods.resTimeHistogram": [250]},
  "tags": {}
}

Total method calls

{
  "appIds": ["YOUR_APP_ID"],
  "type": "summary",
  "metric": "methods.count",
  "aggregation": "sum"
}

Method calls per minute

{
  "appIds": ["YOUR_APP_ID"],
  "type": "summary",
  "metric": "methods.count",
  "aggregation": "rate"
}

Average method response time per host

{
  "appIds": ["YOUR_APP_ID"],
  "type": "timeseries",
  "metric": "methods.total",
  "dimensions": ["tag:host"]
}

The response would look similar to:

{
  "timestamps": [1758600000000, 1758600060000, 1758600000000, 1758600060000],
  "metrics": {"methods.total": [42.5, 38.1, 61.2, 55.7]},
  "tags": {"tag:host": ["web-1", "web-1", "web-2", "web-2"]}
}

List 10 most common errors with when they were last seen

{
  "appIds": ["YOUR_APP_ID"],
  "type": "breakdown",
  "metric": "errors.count",
  "dimensions": ["tag:type", "tag:name"],
  "filters": [{"tagId": "tag:status", "operator": "isNot", "values": ["ignored"]}],
  "extraMetrics": ["errors.lastSeen"],
  "limit": 10
}

Error response

{
  "error": "400",
  "message": "endTime must be greater than startTime"
}