INFRA-XS’s alerting and event management are based on a highly open concept. This means that each element of an alerting definition can be configured freely, including the dispatch mode: besides the “classic” notification to a network management system, individual staff members can be alerted personally – via e-mail or text message.
INFRA-XS provides a number of alerting templates designed to monitor a whole variety of service quality and system parameters: availability and response time behaviour of the monitored workflows or of individual transactions, availability of server systems and monitoring agents, etc. The creation of an alerting definition consists of selecting the suitable template and filling it with data:
The alerting templates come with predefined contents: significant texts – in case of e-mail notification in part supplemented by the corresponding charts graphics – which let you identify the alert-triggering problem at a glance. All the texts can be edited to adjust them to e.g. procedural or operational requirements. In the case of SNMP traps, the adjustment is performed via the MIB variables; besides the default ones dynamic variables can be used.
The basic element of an alerting definition is the threshold, the value of a monitored parameter that triggers an alert. With INFRA-XS there are two possibilities to define a threshold: either use the value defined for e.g. a transaction’s runtime that is stored in the database, or else set a value individually for an alerting definition. This allows the definition of proactive thresholds, alerting you of critical situations well BEFORE they develop into real incidents, and giving you enough time to initiate adequate countermeasures.
As with the monitoring data, the data on triggered alerts are being stored in the database. So you can easily trace the course and development of threshold violations – for instance, to check if taken troubleshooting measures have produced the desired effects.
The notification of alerts can be effected in different ways:
Further reactions which can be triggered are, i.a.:
For the dispatch of alerts via e-mail or text messaging, single recipients or entire groups can be set. To be always up-to-date as to the current state of the alert situation, the dispatch of an end message can be set. In this case you receive a notification as soon as the alert-triggering problem no longer exists.
On the other hand, individual ‘downtimes’ can be set just as well, for example if a staff member doesn’t want to receive e-mail or text message alerts between 10 pm and 6 am – to be all the more able to attend to new SNMP traps or trouble tickets the next morning.
For the event management INFRA-XS applies a dedicated server application: the XS-Problem Server. The server receives the event messages logged on the monitoring agents, processing them according to a set of defined rules. Then the event data, like the monitoring data, gets stored in the database.
The data on events having occurred on the XS-Agents are transferred to the server system, together with the monitoring data, directly after each monitoring cycle. This way, errors and incidents can be detected immediately, and adequate measures can be initiated without delay, like the dispatch of an alert or the generation of a trouble ticket.
There are various possibilities to react to events:
In addition, one of INFRA-XS’s standard reports, provided via the web portal and continually updated, subsumes all the error events related to the monitoring. For deep dive analyses, all types of monitoring events can be utilised, i.e., not only errors and warnings, but also information-type events. This can help you, for example, to deduce patterns in the onset and further development of incidents which you then can define the appropriate alerting rules for.
