Frequently Asked Questions

improve this page | report issue


This topic describes the list of commonly asked questions related to MobileFirst Analytics Server.

In a multi-index elasticsearch cluster, it is important to set the following:

  • Minimum number of shards to be set to the number of nodes in the cluster.
  • Replicas per shard to be set to a minimum of two.

MobileFirst Analytics v8.0 uses multiple indices to store the event data.

In MobileFirst Analytics v8.0, the data store of Elasticsearch has multiple indices. It is not a single index-based data store. Indices are dynamically created based on the type of events flowing into the analytics. So the end users need not be concerned about the multiple indices. Here every index within the Elasticsearch is split into the number of shards set in the config file.

Ensure that the hardware sizing calculator is used for checking the right hardware against the data and the customer requirements. Several factors influence the performance of the system including hardware, the type or size of data events that come into the analytics server and the volume of events.

No. Once the data is purged, it cannot be recovered.

The TTL properties are not applied to the data that exists in the Analytics platform. You must set the TTL properties before you add data.

Ensure that the MobileFirst Server JNDI properties are used to configure the right Analytics end points. Ensure that the Date filter is correctly set for the data to be rendered.

To invoke the Elasticsearch REST APIs, it is mandatory that the property analytics/http.enabled has to be set to true in Analytics server's server.xml.

No. While using IBM WebSphere Application Server Full Profile or Network Deployment (ND), make sure to use the IBM JDK that is provided out of the box with WebSphere Application Server.

First time when the application is opened the App Sessions is zero. When the end user takes the mobile app to the background and brings it back to the foreground, then this action increments the App Sessions to 1. Further repeating the same action continues to increment the App Sessions.

Cluster health YELLOW might not be an issue. Mostly, when there are unassigned shards the cluster health is shown as yellow. When new nodes are joined to the cluster, Elasticsearch reallocates the unassigned shards to the new nodes, thus making the cluster health GREEN. Sometimes, too much of shard counts also leaves the shards unassigned to any of the nodes and hence the cluster health status shows as yellow. Make sure all the nodes in the cluster are active and are working fine and that the shards are in started/active state.

For Web apps, AppSession count is incremented based on the browser session and is based on the connection from the browser (app) to the MFP Server.

Let us say if the browser is using the general window/tab and performs a connection to the server then the App Session count is incremented by one. In the same browser, if the user opens the app on another tab and performs the connect, then the session doesn't increment. The session stays inactive for 30 mins. When you try to reconnect again it is incremented by one.

If the user clears the browser cache and tries to connect, then the device is considered to be a new one and the device count is incremented. Since browsers don't have a real device ID, an ID is generated for the browser app until the offline files/cache is cleared.

This applies to an incognito browser window as well, if you use an incognito browser window and try to connect, an app used to connect from each tab is considered as a new session and the session count is incremented. If the user uses two different browsers and accesses the app to connect to MFP Server, then the device count is incremented by two.

Active users is the number of users using the app. Every unique user is counted as a user using the app. By default deviceID is userID. However app developer can use setUserContext(userid) API. This is going to replace the userID to the value the app developer sets.

One solution/approach is to generate a uniqueID from the computer when the user is accessing the WebApp and it sends that as a customData. This data can be used to compute the statistics for the actual machines (or computers/browsers) from which user accesses the app and uses the setUserContext to set the userID. This data can also be used to generate the custom charts.

In Analytics 8.0, calculating an App session is completely different from any of the previous versions of MFP Analytics.

An App session count is incremented by one when the app is brought to the foreground from background. To enable this for Cordova Apps, we need to enable the CLIENT APP LIFECYCLE events. Please refer here, for more information.

Inclusive terminology note: The Mobile First Platform team is making changes to support the IBM® initiative to replace racially biased and other discriminatory language in our code and content with more inclusive language. While IBM values the use of inclusive language, terms that are outside of IBM's direct influence are sometimes required for the sake of maintaining user understanding. As other industry leaders join IBM in embracing the use of inclusive language, IBM will continue to update the documentation to reflect those changes.
Last modified on May 18, 2018