Use these tasks and information to resolve problems with the Experience Analytics database configuration.
Unable to connect
If you are unable to access the Experience Analytics database, you should verify that you are using the appropriate full-qualified host name.
If you are running against a named database instance, then the host name must follow a specific format:
<host_name> is the IP address, the machine name, or
<instance_name> is the name of the DB instance.
During upgrade, Database Manager fails to upgrade the Result Set Extractor
In some situations, the Database Manager fails to upgrade the Results Set Extractor. The required views do not exist. The RSE Database must be re-created.
Failing to create database in clustered server environment
During installation into a clustered server environment, if you are seeing error messages during creation, you may have improperly specified your SQL path to a disk that is not part of the cluster group or dependency list.
To resolve the cluster dependency issue, see information on the Microsoft Support site about "How to create databases or change disk file locations on a shared cluster drive on which SQL Server was not originally installed" .
In the Database Manager, verify the path where the database files are stored. If you do not know the path to a remote server, contact the appropriate person.
Experience Analytics database performance can be effected by the amount of memory allocated. You might need to adjust memory allocation for Experience Analytics databases to resolve some problems.
If any of the following conditions occur, the memory that is allocated for the SQL Server might be set too low:
- Experience Analytics is performing slowly
- Reports are delayed in arriving
- The box running SQL Server is having paging or disk I/O pro
Verify and change the amount of memory that is allocated to the Experience Analytics databases in the SQL Server Management Studio.
Experience Analytics database performance can be effected by database fragmentation. Normal database operations can cause the hard disk drive media and the stored databases to become fragmented. Over time, disk fragmentation can significantly affect performance.
The Experience Analytics database indexes can also become fragmented. Periodically, these should be defragmented to ensure optimal performance.
Through the Database Manager, you can check the fragmentation status of your database indexes and run defragmentation tasks as needed.
Database Manager hangs when upgrading the System database
The Database Manager can hang when the System database is being upgrade if other components of the system are running.
While this situation can be caused by multiple factors, in many cases the cause is the presence of an active Stats Logger session agent in the Windows™ pipeline of the Report Server. This session agent is used to collect statistical information in the pipeline and submit it to the Statistics database for use in Portal-based reports.
The active session agent might be keeping the Statistics database open and therefore unavailable to the Database Manager, which causes it to hang.
To resolve this issue, try any of the following solutions:
- Stop the Transport Service before you start the Database Manager. Note: As soon as you stop the Transport Service, hits are dropped and data is lost.
After you use the Database Manager, you must remember to restart the Transport Service through TMS.
- Remove or disable the Stats Logger session agent from your processing pipeline. Note: Disable the Stats Logger session agent during Database Manager operations in an All-in-One system.
- When the session agent is removed from the pipeline, you can start the Database
Manager.Note: Remember to add it back through TMS after you exit the Database Manager.
- If the Stats Logger session agent is part of a dedicated child pipeline in the Canister, you can disable the whole pipeline through TMS.
Poor data collection performance
If the data collection process is consuming large volumes of memory or failing to finish, you might need to adjust the Data Collector Batch Size setting, which defines the maximum number of records to extract or load for the Data Collector in a single batch.
Slow data collection times
Typically, data collection should run in under 5 minutes. If the data collection process is more than 10 minutes, then you should increase the size of the Data Trim Interval setting.
Database connection timeouts
If your Experience Analytics databases are experiencing repeated timeouts, you might need to adjust the Database Connection Timeout setting.
The default connection timeout is set to 30 seconds. After the system ran for a while and traffic increased, your hardware might not be able to manage the traffic level at the current setting.
Try setting this value to double its current value. If it happens again, double it again.
Data collection performance from multiple canisters
By default, Experience Analytics can collect from two canisters at one time. If your
environment has more canisters, you can increase the size of the
Data Collection - Max
Concurrent setting. The appropriate setting depends on your hardware environment.
If the size is too large, you might notice any of the following symptoms:
- Slow-down in data collection
- Errors in the data collection process
- The canisters are busy
Decrease the size of the
Data Collection - Max Concurrent setting.
Database is getting too large
If the size of your database is pushing the limits of your storage, you may consider adjusting some settings.
- If you do not use the Page Performance reports, you do not need to collect the Path Statistics. You can disable the Path Statistics collection to save significant database space.
- In SQL Server Management Studio, check the Recovery Model setting for each database. If the
Recovery Model is set to
Full, the databases and their log files can grow larger.
Error - an existing connection was forcibly closed by the remote host
When this error occurs frequently, you should check log information as described here.
From time to time, the Data Collector may register the following log message:
An existing connection was forcibly closed by the remote host
This error indicates that SQL Server or the server hosting it closed the connection unexpectedly. Typically, SQL Server was restarted, or an operation that required closing all existing connections was run, such as a backup or restore operation.
When these errors occur sporadically, they do not affect data collection or the validity of the data.