DRAFT On the Anomaly Impact Deatils page, users gain insight into why & how AGILITY triggered each anomaly notification by delineating the anomaly’s unique information we received. The Anomaly Impact Details page allows you to drill down further into an anomaly's causes & impacts to expedite network resolution processes and gather knowledge to prevent similar anomalies from reoccurring.
To navigate to the Anomaly Impact Details to view the specifics of a particular anomaly:
Information headers at the beginning tell you what you are looking at
Duration is the duration of time that the anomaly lasted
Anomaly Score is back end score we use to determine if the anomaly should be displayed - will be removed
Detected on tells them when we discovered that there is an issue - when our AD model detected it - showing the latency for how long it takes AGILITY to detect that there was an issue. Note that there can be a delay in the receipt of data from the customer which contributes to the latency.
Service data is at the top based on individual user KQIs key quality index reflected by user experience
- toggle button filter data volte
example to use: https://oci-ad-dev.b-yond.com/ran/anomaly/9892917b-580a-e555-c78b-427f1a1bf4b3?restoreFilters=3053639b-3afd-4ed8-a37d-edb44bb954db
KQI name how we ingest the data
QoE = quality of experience : user experience faced on that cell
Value: the output of the KQI. its the aggregated data of individual data of subscribers on those cells. Each user . Value is a measure of the impact of the anomaly.
Sevrice : the service name
Current impact: the model computes the impact and tells you if it is low medium high or not available . No means it is not impacted there are three no, medium, high N/A means the data is not available
See Trends:
cLICK ON THE sEE tRENDS icon to see the Service KQIS Trend Chaart
Auxillary data togle adds secondary data the secondary axix added on the right hand side adds data that lets you visualize the cause
key performance indicators are network level performance level - second two tables PM KPIS
names are standard they all know what those names mean
kpi value mathematical value the result of the anomaly on performance
Imapct Level means the severity of the anomaly and the percentage deviation level how much it deviates from the normal network
Importance : the inidcagti=or that was decided to trigger the anomaly numbers greater than zero indicate that this KPI was degraded and therefor triggered an anomaly . You will see dgredation in the chart if it is higher than 0
FM Alarms
if we have an alarm on site it will trigger an anomaly based on the alarms right now we don't have an alarm . In some cases we do have an alarm. Example of cases with an alarm: https://oci-ad-dev.b-yond.com/ran/anomaly/c3818f35-a987-d578-aa66-58f7bcb3569c?restoreFilters=4bdbf827-75e0-4620-99d1-4b55ce3ebcba
Alarm type: sometimes it is not a KPI issue . Either an issue is a PM KPI or an FM Alarm Fault. Management PM performance Management
Anomlay Correlation : Anomaly correlation between anomalies that are all connected together. Ex there are several towers in a stadium, Tower a had an issue tower b & c had an issue and they were triggered separately, The correlator will bring these issues into one place to advise that here is a common issue happening on an area on a group of towers whether its on the same domain or a different domain ex. a router far away fro m the stadium has an issue and at the same time there was a problem on those sites. if we ave an anomaly in PS core and we notice that it is affecting towers , we connect them
We will get an example from Toufic:
View details will take you to the NOMALYT DETAILS FOR THAT SPECIFIC ANOMALY IT is a link for that particular anomaly
its telling you the anomaly at these two sites is correlated
IMPACTED SUBSCRIBER GROUPS OVERVIEW: change to impacted customer insight
How many users were actually being impacted by that anomaly
SUBSCRIBER GROUPS IMPACTEDCharts help identify who the affected users are general , normal , VIP, corporate government
IMPACTED QoE Qulity of Experience how were the users affected , what happened to them ?
IMPACTED QoEWhat was affected what kind of application was impacted what kind of application they were using social media audio
Deatils column last both volte or data or one . Actionalbel click to see what degradation was faced
