The Live API dashboard provides information on queries sent by Google Flights Search (GFS) to your API endpoint.
The dashboard displays data from the past 7 days and focuses on two key areas:
- Query success rate: Shows the percentage of successful queries, broken down by dimensions.
- Unsuccessful query samples: Provides examples of queries that failed to process, aiding in troubleshooting issues.
Setting the Scope of the Data
By default, you'll see reports for all queries made over the past 7 days in the markets you're enabled for (based on your user country).
You can narrow the information down using the filters on the right side of the dashboard. These filters include:
- User Country (ISO 3166-1 alpha-2): The two-letter country code of the user who made the query.
- Origin Country (ISO 3166-1 alpha-2): The two-letter country code of the departure location.
- Destination Country (ISO 3166-1 alpha-2): The two-letter country code of the arrival location.
- Origin Airport (ATPCO 3 letter code): The three-letter code of the departure airport.
- Destination Airport (ATPCO 3 letter code): The three-letter code of the arrival airport.
- Trip type: Whether the trip is one-way or round-trip. Other trip types are not currently supported by Live API integrations.
- Passengers Adult: The number of adult passengers.
- Passengers Non-Adult: The number of non-adult passengers.
- Preferred Cabin Class: The preferred cabin class selected by the user (e.g., Business Class).
Please note that these filters apply to all reports on the dashboard.
Query results
Summarizes the number of successful and failing queries over time.
Query failure rate
Summarizes the rate of failing queries over time. It is useful to identify time-related failure patterns.
Response time
Shows the 50th, 90th, and 99th percentile latencies over time, measured in milliseconds (e.g., 15000 milliseconds is 15 seconds). A 90th percentile below 15 seconds is generally considered fast enough for users on the Google Flights "Best" tab.
Query results per user country
Provides a view of the query success rate per user country. A low query success rate for a given user country might be a symptom of an issue with a country specific endpoint.
Proportion of queries with no solution
The proportion of successful queries where no solution is parsed over time. A value close to 100% indicates a potential issue with parsing, an outage on your end or a lack of coverage in some markets.
Query results per preferred cabin class
Provides a view of the total queries and query success rate per users’ selected cabin class.
Query results per selected number of passengers
Provides a view of the total queries and query success rate per users’ selected number of passengers, for adults and non-adults.
Query results per trip type
Provides a view of the total queries and query success rate per users’ selected trip type. Currently, Live API supports one-way and round-trip only.
Query results per origin/destination country
Provides a view of the total queries and query success rate per users’ origin, respectively destination, country of the flight. Note that some airports in claimed territories might be reported as ZZ country.
Query results breakdown
Provides a view of the total queries and query success rate for all dimensions (user country, origin country, destination country, origin airport, destination airport, preferred cabin class, number of adult passengers, number of non-adult passengers).
Unsuccessful queries samples
Shows a sample of failed queries. You can download the request and response details for advanced troubleshooting. Sensitive information, like passwords and tokens, is masked with '*' characters.
The following error states are possible:
- TIMEOUT: Your API did not respond within 45 seconds. No response is available for download.
- UNREACHABLE: Your API returned an HTTP error code (e.g., 502).
- ERROR: Your API's response could not be parsed.
Please note that due to the amount of data, only a set of approximately 1000 results are shown.
Common Issues
Load dependent failure pattern:
This occurs when the query error rate increases during peak load periods. This might indicate a potential server sizing issue.
User country depend failure pattern:
If query success rates vary significantly between different user countries, it could indicate a problem with specific country-level endpoints.