Alerts FAQsWe are currently bombarded with alerts. Will netPrefect allow us to filter alerts?
Yes, netPrefect allows you to either not generate the alerts unless specific conditions exist, for example, the alert state/severity has changed, or between certain times of the day, etc. Additionally, in the GUI, you can choose to only display alerts greater that a specific severity, or after a specified date/time, etc. Filtering alerts is completely tailorable to make sure that you only get the alerts that are really meaningful. Would it notify if backup had failed? We currently use Backup Exec.
Yes. As with all event detection, netPrefect takes the message from your application (in this case Backup Exec), which has an indication of job status. From this message we will be able to automatically determine success or failure and notify you accordingly depending on the rule you set. Our main concern is that we get an alert as soon as the disk drive starts creating errors, whether they’re soft errors or hard errors, and what we don’t want is to get the errors when the drive is actually down. How does netPrefect cope with this?
The regular polling for device status should indicate the current status of the device, which may include hard and soft errors. netPrefect can be configured to alert if the error count increases, and can also indicate what replacement part number is required on site when an engineer visits. The key to netPrefect is its proactive alerting. My customers are demanding a better service. How can netPrefect help?
netPrefect allows you to be informed of issues as they arise instead of when the customer notices and calls you. In some cases, it is also possible to automatically correct the problem by executing commands on a managed device indicating its status to you. Can you filter alerts by time of day?
Yes, alerts may be sent to different destinations (email, GUI, etc.) depending on the time of day or day of week. Can I stop alerts when I have a system down for maintenance?
Yes, simply use the netPrefect GUI to set the device (or subsystem) into Maintenance mode with an optional time at which the device should be placed back into production status. Does the dashboard provide a simple graphical notification of alerts such as using individual icons for clients, servers and services?
The dashboard view in the main netPrefect GUI simply indicates the total number of alerts that are present, categorised either by severity or classification. The web interface can display more information to the user, such as uptime pie charts and time series information, detailing any of the statistics gathered by netPrefect. What about checking SQL backups? There is an event when backup is done, so if you don’t get that, can netPrefect send a message?
Yes, netPrefect can generate an alert when something of interest doesn’t happen (such as a backup not running). Will it tell me the part number? That would be really useful.
netPrefect can be configured to look up part numbers for a device from an external database and then notify engineers of the part number. The alert can also contain expert knowledge about the device or the alert itself – anything that might be useful to advise an untrained operator. Can you set netPrefect to say if it’s not seen an event in the event log in the last 2 hours? What we would be looking for in the event log would be a successful message within the last 5 hours.
Yes, netPrefect can automatically raise an event if a condition has not been seen for a specified length of time. This event may then cause an alert to be raised depending on the rules you have set. How does netPrefect let me use an individual alert to drill down to find out exactly what the problem is?
netPrefect displays the outstanding alerts in a list form. In the case of alerts that have been propagated up the hierarchy of managed devices, these alerts can be shown in more detail, including the alerts that are their root cause. This can be annotated with any information that the developer of the rules that generate the alerts has included, which can be particularly useful for problem resolution. When you investigate an alert, can it all be conducted through a single screen?
Yes, all interaction with netPrefect and the devices it is managing is performed through a single GUI (each user has their own instance of the GUI, with their own privileges). If just the ability to view details about a managed device is required, then a web browser can be used to display them directly without the need to install the GUI. This is ideal for giving your customers the ability to view without interaction with the device. If a service has stalled and we then restart it, will netPrefect identify whether it’s a hardware failure so that we can get a service engineer in?
If the information detailing the root cause of the service fault is available to netPrefect (either in log files, or via SNMP/WMI queries) then yes, netPrefect can determine the hardware that the engineer needs to replace and, if configured correctly, can even indicate to the engineer the part numbers and locations to get them. If the server power dies, will netPrefect identify and report this?
netPrefect will detect that the server is no longer functioning. However, it will be unable to determine the cause, unless the server is connected to a UPS that netPrefect is also monitoring. Can netPrefect set thresholds?
netPrefect can be configured with thresholds on key performance statistics. These thresholds, when crossed, can cause action to be taken, such as sending an email or generating an alert. Does netPrefect allow us to check SQL databases up and online, exchange servers up and online and any excessive queuing with exchange server?
Yes, using the WMI transport, all performance characteristics of SQL and Exchange can be monitored. If I correct an issue that has caused an alert, do I have to wait until the next scheduled time for the alert to be removed?
No, you can either manually re-issue the scheduled query or, in the case of the reachability transport, which has no interactive console, it can be reset which will force an immediate check for reachability. |
|
|
|
