To ensure optimal uptime for app access and connectivity, you should monitor the following core components in the XenMobile environment:
XenMobile server generates and stores logs on local storage that you can also export to a systems log (syslogs) server. You can configure log settings to specify size constraints, log level, or you can create custom loggers to filter specific events. You can look at XenMobile server logs from the XenMobile console at any time. You can also export information in the logs via the syslog server to your production Splunk logging servers.
The following list describes the different types of log files available in XenMobile:
Debug log file: Contains debug level information about core web services of XenMobile, including error messages and server-related actions.
<date> <timestamp> <loglevel> <class name (including the package)> - <id> <log message>
- where <id> is a unique identifier like sessionID.
- where <log message> is the message supplied by the application.
Admin audit log file - Contains audit information about activity on the XenMobile console.
Note: The same format is used for both admin audit and user audit logs.
With the exception of required Date and Timestamp values, all other attributes are optional. Optional fields are represented with " " in the message.
<date> <timestamp> "<username/id>" "<sessionid>" "<deviceid>" "<clientip>" "<action>" "<status>"
"<application name>" "<app user id>" "<user agent>" "<details>"
The following table lists the available admin audit log events:
User audit log file: Contains information related to the user activity from enrolled devices.
Note: The same format is used for both user audit and admin audit logs.
With the exception of required Date and Timestamp values, all other attributes are optional. Optional fields are represented with " " in the message. For example,
<date> <timestamp> " <username/id>" "<sessionid>" "<deviceid>" "<clientip>" "<action>" "<status>" " <application name>" "<app user id>" "<user agent>" "<details>"
The following table lists the available user audit log events:
NetScaler also monitors the XenMobile web service state, which is configured with intelligent monitoring probes to simulate HTTP requests to each XenMobile server cluster node. The probes determine whether the service is online and then respond based on the response received. In the event that a node does not respond as expected, NetScaler marks the server as down. In addition, NetScaler takes the node out of the load-balancing pool and logs the event for use in generating alerts through the NetScaler monitoring solution.
You can also use standard hypervisor monitoring tools to monitor the XenMobile virtual machines and to provide relevant alerts regarding CPU, memory, and storage utilization metrics.
SQL Server and database
SQL Server and database performance directly affects XenMobile services. The XenMobile instance requires access to the database at all times and goes offline (for example, stops responding) in the event of an outage to the SQL infrastructure. The XenMobile console may continue to function for a while following any disk space issues with SQL Server. To ensure maximum database uptime and adequate performance for the XenMobile workload, you should proactively monitor the state of your SQL Servers following Microsoft recommendations. Additionally, you should adjust resource allocation for CPU, memory, and storage to guarantee service level agreements as your XenMobile environment continues to grow.
NetScaler provides the ability to log metrics to internal storage or to send logs to an external logging server. You can configure the syslog server to export NetScaler logs to your production Splunk logging servers. The following logging levels are available in NetScaler:
The log files are also stored in NetScaler storage in the /var/log/ns.log directory and named newnslog. NetScaler rolls over and compresses the files by using the GZIP algorithm. Log file names are newnslog.xx.gz, where xx represents a running number.
NetScaler also supports SNMP traps and alerts as a monitoring option. The following table lists the SNMP traps recommended for the environment.