Pre-launch Checklist

Where to manage your Client ID in the Google Cloud Console

The Premium Plan Client ID management functionality is migrating from the Support Portal to the Cloud Console on the Maps' Credentials page, under the Service Accounts section.

The new Client ID area on the Credentials page

Note: The Google Maps Platform Premium Plan is no longer available for sign up or new customers.

Ensuring your team has access to the necessary resources

Keep your Google Maps Platform Premium Plan welcome letter somewhere safe

Why it's important: Your welcome letter is your Google Maps Platform Premium Plan starter kit and perhaps also your first aid kit. It contains critical bits of information such as your Google Cloud Console project ID, your client ID and your cryptographic key, which are necessary to begin using the Premium Plan. It also contains all the information you need to contact the Premium Plan Support team if you experience any technical issues with any of the Google Maps APIs.

Use the Google Cloud Console

Why it's important: The Google Cloud Console gives you access to information such as usage reports, news feeds, and developer resources. More importantly, the Cloud Console allows you to file support cases with the Premium Plan Support team if you encounter any technical issues during development or launch.

Prior to launch, please enable Cloud Console access for all developers responsible for your application's maintenance. If you experience technical issues, access to the Cloud Console will allow members of your team to contact support and also let our support team to contact the proper stakeholders in your organization. For example, the support team may need to contact your organization if we detect abnormal traffic or behavior that could end up breaking your application. Ensuring that we can contact the appropriate developers could be the difference between having an unexpected outage and preventing an outage.

Subscribe to notification email groups

Why it's important: To ensure you stay up to date with developments and changes across the Maps APIs, we recommend subscribing to one or more of the following email groups:

Subscribe to relevant notification feeds

Why it's important: To ensure you stay up to date with developments and changes across the Maps APIs, we recommend subscribing to the relevant notification feeds, as described in the FAQ.

You can also subscribe to the following RSS feed for Google Maps Premier API Announcements: Outages, Updates, Service Notifications:

http://google.force.com/services/xml/MapsRSS

Have the support hotline handy

1-877-355-5787 for US customers, +1 404-978-9282 for customers outside the US

Why it's important: The hotline is your way to phone support. The numbers for each country are available in the Cloud Console, along with the PIN. While you are welcome to use the support hotline to report technical issues to our team, it is reserved for production-down, service-unusable cases only. Our priority levels are defined in this document.

Optimizing your application

Configure a firewall to allow access to the Google Maps Platform Services

Why it's important: Google Maps Platform services use a variety of domains, some of which do not belong to the *google.com domain. If you are behind a restrictive firewall, it is important to allow access to the domains used by each Maps API service. If your firewall doesn't allow access to these domains, API requests will fail, which can break your applications. See a complete list of domains used by the Maps APIs.

We do not recommend managing firewall restrictions by IP address, as the IPs associated with these domains are not static.

Note: Google Maps Platform services use port 80 (http) and 443 (https) for inbound and outbound traffic. These services also require GET, POST, PUT, DELETE, and HEAD requests. Configure your firewall to allow traffic over these ports and to allow requests, depending on API and use case.

Load the APIs over the correct SSL hostname

Why it's important: Applications that load the Maps APIs over SSL should do so from https://maps.googleapis.com rather than the legacy hostname, https://maps-api-ssl.google.com.

Authorize your SSL domains for use with the Maps JavaScript API

Why it's important: When using the Maps JavaScript API with an SSL domain, it's critical you have explicitly authorized your HTTPS domains to ensure your requests are not rejected. Note that authorizing http://yourdomain.com does not automatically enable its SSL equivalent, https://yourdomain.com. Check your list of authorized domains in the Cloud Console by scrolling down to the Client ID section. To troubleshoot errors related to using the client-side APIs with an SSL domain, check if any elements of your page are loaded over HTTP. View the guide to troubleshooting authorization.

Select the proper API version

Why it's important: Before developing your application, it's important to be aware which versions of the APIs are deprecated. Choosing to develop against the non-deprecated versions of the APIs will save you development time and cost down the road once deprecated versions become unavailable.

In particular, it's critical to understand the versioning scheme used by the Maps JavaScript API, so that you avoid accidentally using an improper version of the API in your environment.

For example, it may be suitable to use the experimental version of the API in your development or test environment, but we strongly discourage the use of the experimental version in a production environment. Our SLA only applies to stable versions of the API, so you should only use stable versions in your production environment.

See the guide to Maps JavaScript API versions.

Choose between client-side and server-side design

Why it's important: Choosing a client-side or server-side approach is an architectural decision and is absolutely critical to the stability and scalability of your application. By and large, a server-side approach should be used for pre- and post- processing of records offline (that is, outside of your application). Alternatively, a client-side approach should be used for the portions of your applications which interact with your users (that is, process user-submitted requests in real time).

Deploying a server-side approach where a client-side approach should be used instead is the leading cause for exceeding quotas and, hence, broken applications. We highly recommend consulting the geocoding strategies before designing or launching applications that rely on server-side calls.

Optimize quota usage

Why it's important: Understanding the way your application consumes quota, known as Maps APIs Credits, helps you to reduce the amount you pay. For example, if you're using the Maps JavaScript API, your application consumes Maps APIs Credits for each map load. See the guide to Premium Plan usage rates and limits.

Manage your web services quota usage

Why it's important: By default, the shared web services quota is set at 100,000 daily free requests. For a more detailed quota breakdown per API, see the usage limits documentation. Check your quotas in the Cloud Console, or file a support case if you have any quota issues.

Before launching your service, it is critical that you understand the different quota related errors (for example, OVER_QUERY_LIMIT, User Rate Limit Exceeded), and set up the proper logic in your application to be able to respond to such errors when you exceed your quota. Please start by reading the usage limits FAQ. For information on the status codes returned by each API, consult the developer's guide for that API. For example, see the guide to Directions API status codes. Understanding and implementing these concepts will greatly reduce the chances of your application exceeding its allowed quota, being blocked by Google, and/or breaking.

Perform load testing on your app

Why it's important: Use load testing of your application to ensure it can handle high volumes of requests without exceeding your quotas for the Maps APIs.

Load testing against live Google services will lead to your application exceeding its allowed quota and being blocked by Google. Google Maps Platform can serve very high volumes. In 2012, Santa Tracker served 1,600,000 requests per second. Therefore, there is no need to do load testing against Google services. Instead, load testing your application should ensure that your application is able to cope with high volumes of requests without exceeding your quotas for the Maps APIs. Example: if your quota for the Geocoding API is 20 QPS (queries per second), load testing your application should ensure that your application can handle 600 QPS without sending more than 20 QPS to the Geocoding API.

To safely achieve this, load testing should be done against a mock (fake) API—a service that can absorb high amounts of requests and reply to them with valid responses, without involving the Google Maps Platform. Thus you can load test your application without risking being blocked by the Google Maps Platform.

Please see this example of a mock API, implemented as a small Google App Engine application. You can upload this example to your own App Engine application (after you register one at appengine.google.com) and make your application send requests there instead of to maps.googleapis.com.

Default (free) App Engine quotas should generally be enough to load test your application well beyond your quotas for the Maps APIs web services. Please make sure your application sets the correct User-Agent header to enable response compression. This is critical to ensure efficient usage of bandwidth, which is particularly important for an App Engine application serving a high volume of plain text (JSON/XML) responses. If you need higher quota for your App Engine application, you can also enable billing, although this should rarely be necessary.

Migrating your application from a standard to a premium license

Include your client ID or API key in API requests

Why it's important: One of the most important things you can do for your application is to ensure you include your client ID (gme-yourclientid) or your API key (which looks something like this: AIzaSyBdVl-cTICSwYKrZ95SuvNw7dbMuDt1KG0) in your API requests. The client ID or API key identifies your requests as a Google Maps Platform Premium Plan request.

You must include your client ID or API key in your applications in order to benefit from any features specific to Premium Plan. Inclusion of your client ID or API key is also necessary in order to receive technical support and to ensure your application is under our SLA.

For most of the APIs, you can choose whether to use a client ID or API key. Your client ID is included in the welcome letter which was issued to your organization's primary contact(s). You can generate your own API key or keys on the Google Cloud Console.

Details are in the guide to authentication and authorization.

Include either the API key or the client ID in API requests, but not both

Why it's important: In order to correctly load the Maps JavaScript API, or to send a request to other Google Maps APIs, you must include either your client ID or your API key, but not both. If you choose to use a client ID, you must remove any key parameters. If your request includes both a client ID and a key, your application may experience unexpected behavior or errors.

Follow the guide to authentication and authorization for full information on how to correctly format Premium Plan requests per API.

When using a client ID, authorize your domains for use with the Maps JavaScript API

Why it's important: To prevent unauthorized sites from using your client ID, the Maps JavaScript API requires that you authorize all domains with our support team for all sites which will use your client ID. (URL registration is not required if you use an API key instead of a client ID.) If there is no match between the URLs authorized to use your client ID and the site trying to use your client ID, your site will be unable to use the API with your client ID. You can authorize domains at any time so please ensure that you have authorized the domains for all your sites in advance of your launch.

You can check your list of authorized domains in the Cloud Console by going to the Credentials page and scrolling to the Client ID section.

For authorization issues, we recommend that you review the guide to troubleshooting authorization before filing a case.

When using a client ID, sign web service requests using a signature generated with your private cryptographic key

Why it's important: Your private, cryptographic key is used to generate digital signatures which communicate to Google that your requests have come from a trusted source. Our web service APIs require that you add a digital signature to your requests, if you use a client ID for authentication. This adds a layer of security on top of your request which will better safeguard the quota associated with your client ID. Your cryptographic key (for example, vNIXE0xscrmjlyV-12Nj_BvUPaw=) is included in your welcome letter which was issued to your organization's primary contact(s).

Note: The cryptographic key is used to generate signatures. Do not append it to your requests as a signature itself. Your cryptographic key is similar to an ATM pin number. It is used as a means of authentication in order to access your account and should never be openly shared with or visible to untrusted sources. Premium Plan web service requests which are not properly signed will be rejected by our servers so it's critical your application properly signs request prior to launch. See the guide to authentication and authorization.

Track application usage

Why it's important: As a Premium Plan customer, you have access to detailed reports on your application's usage, including requests made, credits consumed, errors returned, and more. See the guide to reports.

A channel parameter is an optional parameter that allows you to track usage under your client ID by assigning a distinct channel to each of your applications. Channel parameters do not need to be registered to your client ID. By adding the channel parameter to your API request, the usage results per channel will begin to appear in your support portal usage reports 1-2 days after implementation. It is up to you to decide where your channels are implemented and, hence, how your usage is aggregated. Please decide prior to launch if your application should integrate channel parameters to track your application usage.

The channel parameter must use the following format:

  • Must be an ASCII alphanumeric string.
  • Period (.), underscore (_) and hyphen (-) characters are allowed.
  • The channel parameter is case-insensitive; upper-case, mixed-case, and lower-cased channel parameters are merged into their lower-case equivalent. For example, usage on the CUSTOMER channel is combined with the usage on the customer channel.

You may implement up to 2,000 distinct channels per client ID.

To use the channel parameter, include it in the request URL together with the client parameter used for passing the client ID.

Please note that the channel parameter must be a statically assigned value per application. It must not be generated dynamically and used to track individual users.