Hello Dojo community,
I am working on transitioning our operations from reactive analytics to a proactive FDIR (Fault Detection, Isolation, and Recovery) model. To support this, I want to automate a weekly export of alarm events/Event IDs. To this end, I’m working on an automation where I want to extract alarm events and event IDs over a defined time period from the Alarm Console to enable proactive FDIR and weekly trending reports. I aim to configure a filter so that the output can automatically be delivered to operators as dedicated report/analysis/analytics
Specifically:
• Is there a way to use wildcard or regular expression syntax in the Alarm Console filter on history alarms (time-bounded view) to capture specific event IDs or ranges especially when there are many ranges to consider?
• I have built a filter around regex range of interest in the past before.*(?:XY0\d{2}|XY100).*, but this is quite different and generic as there are different ranges for different network elements and we will be looking at 3-4 ranges here (find attached the regex as fetched from the console)
• My goal is to avoid having to script the filter externally and to leverage the Alarm Console’s native filtering where possible. For a weekly delivery to operators, would you recommend building this around a GQI (Generic Query Interface) query, or should I look into an Automation Script that queries the database directly?
I’m looking for the most scalable way to ensure these issues are passed through our FDIR pipeline consistently. Any steer or examples of similar implementations, syntax guides, or community experiences would be greatly appreciated!
Thank you!
Hi Samson,
For the daily FDIR, could it work for your use case to use a sliding window to retrieve alarms within a defined time frame, and then apply additional filters (including Regex expressions) to further refine the output? The following example, retrieve the last 24 hours of active alarms and filter them out by element name (different ways to achieve this):
For the weekly delivery, I think a GQI ad-hoc data source is a good option because it will enable the extraction and reporting of the events using complex criteria. We have in our catalog a sample of this GQI that will probably be a good starter for your solution. Find the item here.