A DevOps engineer is troubleshooting intermittent tail latency spikes for an API service deployed across multiple Compute Engine VM instances in the us-central1 region. The engineer needs to configure a chart in Cloud Monitoring Metrics Explorer that meets the following requirements: Scope the metric data exclusively to VM instances located in any us-central1 zone. Restrict the data stream to requests handled by the production deployment tagged with the metric label environment set to production. Group the time series by zone and chart the 99th percentile response latency across instances within each zone using a 5-minute alignment interval. How should the engineer configure the query parameters in Metrics Explorer?

Configure two filter criteria using logical AND: `zone =~ "us-central1-.*"` and `environment = "production"`. In the Aggregation section, select the function "99th percentile", select "zone" in the group-by field, and set the alignment interval to 5 minutes.: ### What this configuration does This configuration applies multi-dimensional filtering, statistical alignment, and cross-series grouping directly within the **Cloud Monitoring Metrics Explorer** interface to isolate and analyze high-percentile tail latency per zone. ### How it meets the requirements - **Targeting all `us-central1` zones**: Metrics Explorer supports **RE2 regular expressions** via the `=~` comparator. Specifying `zone =~ "us-central1-.*"` matches all zones within the region (e.g., `us-central1-a`, `us-central1-b`, `us-central1-c`, `us-central1-f`). - **Restricting to production data**: Adding an exact match filter `environment = "production"` refines the query. When multiple filter criteria are provided in Metrics Explorer, they are automatically joined with a **logical AND**, meaning only time series matching both criteria are included. - **Grouping and 99th percentile aggregation**: In the **Aggregation** element, setting the aggregation function to **99th percentile** derives the statistical 99th percentile across time series within the aligned 5-minute interval, while selecting **zone** in the group-by menu produces exactly one time series per unique zone. ### Key advantages - **Isolates tail latencies**: The 99th percentile captures outliers that would otherwise be hidden by average or mean calculations. - **Reduces visual clutter**: Grouping by `zone` reduces high-cardinality instance-level time series into clean, actionable regional zonal baselines. - **Flexible regex matching**: RE2 pattern matching eliminates the need to create separate manual queries for each individual zone. ### Why this is optimal This approach directly leverages Cloud Monitoring's native reduction and alignment pipeline without requiring custom client-side processing, allowing the engineer to spot zonal anomalies in tail latency rapidly.

Metrics Explorer Filter Aggregation Percentile Query | Professional Cloud DevOps Engineer

Metrics Explorer Filter Aggregation Percentile Query

Worked example. The correct answer is already marked and every option is explained below, so there is nothing to select here. To answer questions yourself, start the free trial.

A DevOps engineer is troubleshooting intermittent tail latency spikes for an API service deployed across multiple Compute Engine VM instances in the us-central1 region.

The engineer needs to configure a chart in Cloud Monitoring Metrics Explorer that meets the following requirements:

Scope the metric data exclusively to VM instances located in any us-central1 zone.
Restrict the data stream to requests handled by the production deployment tagged with the metric label environment set to production.
Group the time series by zone and chart the 99th percentile response latency across instances within each zone using a 5-minute alignment interval.

How should the engineer configure the query parameters in Metrics Explorer?

Configure a single resource filter zone = "us-central1". In the Aggregation section, select "Sum", add "environment" to the group-by field, and set the alignment interval to 5 minutes.

Configure two separate filter criteria using logical OR: zone = starts_with("us-central1") or environment = "production". In the Aggregation section, select "Mean", leave the group-by field empty, and set the interval to 5 minutes.

Configure two filter criteria using logical AND: zone =~ "us-central1-.*" and environment = "production". In the Aggregation section, select the function "99th percentile", select "zone" in the group-by field, and set the alignment interval to 5 minutes.

Set the Aggregation element to "Unaggregated" with group-by set to "None". In the Filter section, apply zone !=~ "^us.*" and environment != "production".

Professional Cloud DevOps Engineer

Metrics Explorer Filter Aggregation Percentile Query

Related questions