- Control of your applications
This application note gives an overview of how to monitor and identify problems directly related to your central applications using op5 Monitor and op5 Log Server. A basic understanding of the actual challenge is the key to successful and effective application monitoring and can be summarized in one question:
“Why is the application X not performing according to our business needs / the promised performance?” (this is a scale from a 100% down application to a situation with unacceptable response time or when incorrect data is produced).
Introduction
Trying to list the applications that cause the most problems is an impossible task and all organizations have their own lists of “key applications” to support their business needs. However, taking a pure quality uptime perspective makes this task greatly simpler. The core challenge of increasing uptime is to limit downtime. To limit downtime we need to look into the details:
According to Gartner, at least 80% of the time fixing a problem is focused on identifying where the actual core problem is. Once we have identified where and what the problem is, the solution is often quite simple and fast to fix.
Modern IP based applications vs. old legacy system applications
In the old days applications were confined to one or multiple similar closed systems. This created a need for extremely detailed information on the particular single application. It was common practice to monitor hundreds or even thousands of highly specific data in each application and by that assuring 100% control of the single system.
Looking at today’s IP based applications they are fundamentally different, both in there architecture and in the design in which they are implemented. The applications are built up by many sub applications and together they provide the organization with the IT systems needed.
To prove our claim is simple – take any modern system from any vendor today and install it on a single system without connecting it to anything. We can almost guarantee full uptime for many years to come! The problem is that you will have no use of such an “off line” system – it needs to be connected to many of your other systems - and connected with quality!
Addressing the challenge from a two way approach
op5 Monitor uses multiple parallel monitoring to assure performance and functionality on your applications. They can be divided into four levels:
1. Check that the underlying hardware/infrastructure is in normal operational status.
2. Check that the included processes in the applications are functioning normally.
3. Do specific checks on the functional status of the different programs.
4. Do generic process checks that will show you the overall performance of a complete function.
1 - Hardware/infrastructure checks
A slow fan, an almost full disk, a broken RAM, a congested LAN or WAN connection will all greatly impact on any applications’ performance and function.
2 – Included processes
Simply checking that IMAP process, the POP3 process, the NDS Process etc. are operational and running is a must to know that the general application will work properly. This applies to both the system and user perspectives.
3 – Specific Checks
op5 Monitor has a rich set of specific application checks for different purposes. In general they all have a common purpose of initiating an action/query>check the response>measure the overall transaction time. Examples of these specific checks are for Databases (MSSQL, MYSQL, Oracle etc), Citrix, Email /Exchange etc. For a complete list please contact us.
4 – End to end service/application checks
As a complementary way to assure full quality we can also create complete standard based scripts for a fixed process containing multiple steps. As an example: testing an extranet order application the op5 Monitor can access the extranet with a username and password, get confirmation that the authentication system is operational, place a dummy order, get the order acknowledgment back, read the order to see that it is OK and finally time the complete transaction. This test is executed every 3 minutes and will tell you that the application is functional from a 100% end user perspective.
The “Business critical application” is on purpose drawn as a cloud as it is a mesh of different objects that together builds the overall application. Our key argument when monitoring applications is the need to check and measure the overall performance from all four levels to assure quality and uptime for the services.
In order to reduce downtime it is imperative to be able to see the complete picture. This enables you to fast identify the root cause of a problem when a service or application is malfunctioning.
op5 Service group view of the generic “Mail service”
In order to secure quality for this Email service, op5 Monitor checks all levels:
Any errors or quality degradation will automatically generate an alarm and clearly show if it affects the general application Email.
Automatic SLA follow up and reports
A direct control of your application performance in a specific SLA/uptime fashion is directly available over a simple web access. These reports can further be configured to be automatically generated as weekly or monthly, and distributed directly from the op5 system via email keeping management informed. Read more about SLA monitoring and available reports in op5 Monitor.
Adding op5 LogServer for even more detailed application monitoring
Logs in general are often considered to be the most granulate source of detailed data. The best part is that it is already there and in your IT systems today. The worst part is that it contains so much data and is distributed over a wide range of devices to make it almost impossible to use for day to day application monitoring.
Applying op5 LogServer addresses key challenges:
- Centralizes all logs from all environments into one central storage
- Enables smart and easy to use filter management
- Takes huge amounts of data and quickly turn it in to usable and valuable input for your application monitoring.
The op5 LogServer handles all logs from applications, hardware and traffic devices. That way it enables the possibility for extremely detailed views of an incident from different angles at all levels of the IT infrastructure.
Automatic Alarms from Logs
Using smart filters in the op5 LogServer its possible to identify specific events on this very detailed level and then trigger automatic alarm notification in the op5 Monitor system. The op5 Monitor system then notifies the responsible person or group about the incident based upon the logs.
Another great benifit from using logs to trace and monitor applications is that you will never be “out of date” as the log information (i.e. the actual information sent in the logs) is always automatically upgraded when your system is upgraded.
Summary
op5 Monitor adds great value by easily and effectively delivering the most important part of monitoring – the overall view, enabling fast root cause identification. Adding op5 LogServer to the application monitoring solution enables quick and very granulate data to the quick and accurate proactive monitoring. It is all done using open standard based formats and tools.
For further information please contact us on +46-8-58 83 01 00 or info@op5.com
|
