Integrate Network State into Applications Manager

Applications Manager | February 8, 2010 | 2 min read
Today’s applications are complex in nature. Applications are hosted in data centers, thousands of miles away from Head Office / Branch Offices. Administrators / Helpdesk Technicians face an uphill task to troubleshoot Application performance or issues, though they have sophisticated tools to remotely access these applications. Users frequently post these questions to Administrators:

  1. I am not able to access the application, though I am able to ping the server.
  2. The application is very slow.
  3. I am able to login to the application, but the page takes more than a minute to load

Technicians, generally, don’t know where to troubleshoot this issue. The Business Service seems to look normal when the Technician uses it. But, who owns the responsibility for the incident reported by the user? This usually is a million dollar question, as no one would want to own responsibility for any application related issue – be it the Network Team, or the Server Team, or the Database Team, or the Application Team.

Let us find an answer to this question through the Network Connector Add-On in ManageEngine Applications Manager with an example. JBOSS Application, with an Microsoft SQL backend database is hosted in Data Center in US. The Business Service is accessed by Users from the Head Office in India.

The screenshot below shows how the application server has been performing.

The health and availability of the JBOSS Application is clear, which indicates that the Business Service has no issues at the application end.

Let us look at the Database server.

The availability of the MSSQL database server is clear. However, the health is critical. If you take a closer look at the Alarm message, you can see that a couple of databases are offline and a Scheduled job is critical. These incidents cannot be deemed responsible for the application to perform slowly. So, let us ignore these alarm messages and find out if there are any spikes in the performance of the server

The performance parameters such as CPU, Memory and Disk are well within Threshold limits. The next option to look at is the Network Devices.

It can be inferred that the health of the Network devices has affected the performance of the Business Service. A Business View connecting the Data Center and Head Office from OpManager can be included into Applications Manager Dashboard. The Business View, as shown below, indicates that the link between the Data Center and Head Office is critical, which needs immediate attention.

The slow performance of the Business service may be attributed to any one of the reasons below :
 

  1. The link between these offices may be down
  2. The RTT is beyond the threshold limit

The technician can now conclude that the application was performing slowly due to the link failure between the Data Center and Head Office. He can escalate this issue to the Network Team. The Network Team can use ManageEngine OpManager to find out the root cause of this issue.

This capability of ManageEngine Applications Manager is in addition to the SNMP & Ping Monitoring. By using ManageEngine OpManager, you can get out-of-the-box support for Network Devices.

To know how to integrate Network Devices from OpManager into Applications Manager, read the guide @ http://www.manageengine.com/products/applications_manager/help/monitors/network-application-monitoring.html

You can follow my post “How to connect Network with Applications? ” for the detailed steps to configure the Network Add-On in Applications Manager.

Kevin

thaveedu