cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
822
Views
10
Helpful
5
Replies

WxCC Queue Stats SLA from API

Sascha Monteiro
Level 6
Level 6

Hi,

I am getting the queue stats from /v1/queues/statistics for the current day, but the serviceLevelThresholdPercentage is often 0.0,

I just tested a couple calls where I answered one after 30sec (sla is set to 20sec), the API shows 0.0, but the analyzer dashboard shows 80%

 

Any ideas?

 

I did see some values yesterday, not sure why

 

This is the current output:

[{"intervalStartTime":1648407600000,"queueId":"xxx","queueName":"Support","channelType":"telephony","totalOfferedTasks":4,"totalEnqueuedTasks":4,"totalAssignedTasks":0,"totalAcceptedTasks":4,"totalRejectedTasks":0,"totalAbandonedTasks":0,"averageEnqueuedTime":3961.5,"averageHandledTime":29592.5,"serviceLevelThresholdPercentage":0.0},{"intervalStartTime":1648411200000,"queueId":"xxx","queueName":"Support","channelType":"telephony","totalOfferedTasks":2,"totalEnqueuedTasks":2,"totalAssignedTasks":0,"totalAcceptedTasks":1,"totalRejectedTasks":1,"totalAbandonedTasks":0,"averageEnqueuedTime":15161.0,"averageHandledTime":9930.0,"serviceLevelThresholdPercentage":0.0}]
1 Accepted Solution

Accepted Solutions

Arunabh Bhattacharjee
Cisco Employee
Cisco Employee

Thanks for the additional details @Sascha Monteiro, very helpful.

It seems like we faced an issue on production related to a recent deploy, the cisco incident ID was CJPIM-632. It affected EU,UK and ANZ datacenters though, was not supposed to affect US (where I tested).

Could you please confirm your datacenter / api URL where you are retrieving the stats? e.g mine is api.wxcc-us1.cisco.com 

It is highly likely that you hit the issue.

The deployment had since been rolled back so population of SL% was resolved. Since the stats package (data service) was deployed and rolled back the metrics are likely to be off until corrective scripts have been run. The ETA for those is over the next few days.

 

A formal RCA will be released so you can reach out to your CSM/PSM from Cisco for more information on this by tagging the incident ID with them.

 

Thanks & Regards,

Arunabh.

 

View solution in original post

5 Replies 5

Arunabh Bhattacharjee
Cisco Employee
Cisco Employee

Hello Sascha,

 

Thanks for reporting your observation.

So, the serviceLevelThresholdPercentage from the definition (reviewed from the API spec) - is supposed to give the % of the accepted calls (by an agent) that have breached the serviceLevel that is defined.

 

Here is the formula I found was being used by this API:

 

serviceLevelThresholdPercentage is 0.0 by default, only if there are any calls that have breached SLA during that period/interval, and only if ended calls > 0 is this value calculated for that interval. Otherwise it is 0.0.

 

So, if I sent 1 call into the queue, accepted and ended that call, and that call breached service level, I will get 100.0 in that interval.

 

i.e 

(endedCallsCount - slaBreachedCallsCount) / endedCallsCount * 100

 

I have tested this and it works as per the formula defined but note that the SL level can be defined both at the Entry Point configuration and Queue configuration so during testing you may want to align the settings.

 

Also note that the metrics for queue stats gets updated in non-realtime - it is not instant, approx 3 mins, there is an enhancement to make this more realtime (the computation job runs approx N mins and calculates these for the completed tasks).

 

Let me know if you are able to test these scenarios and find any anomalies with this. 

I have attached screenshots from testing (50%) service level percentage, it was applied as 30s and I answered one call within and one call outside (breached)

 

Thanks & Regards,

Arunabh.

thanks, it stayed 0 all day and I had a call breaching SLA, the wxcc analyzer dashboard showed 80% SLA when there were 5 calls and 1 of them was answered over SLA, but the API still showed 0.0 for all intervals yesterday even a few hours after those calls...but today it appears to have values. When I run the query with yesterday's start/end timestamp it still shows 0.0 for all intervals

if you get 1 call and sla is breached for the call you should get 0 (%)...right?

1 - 1 / 1 * 100 = 0

Arunabh Bhattacharjee
Cisco Employee
Cisco Employee

Thanks for the additional details @Sascha Monteiro, very helpful.

It seems like we faced an issue on production related to a recent deploy, the cisco incident ID was CJPIM-632. It affected EU,UK and ANZ datacenters though, was not supposed to affect US (where I tested).

Could you please confirm your datacenter / api URL where you are retrieving the stats? e.g mine is api.wxcc-us1.cisco.com 

It is highly likely that you hit the issue.

The deployment had since been rolled back so population of SL% was resolved. Since the stats package (data service) was deployed and rolled back the metrics are likely to be off until corrective scripts have been run. The ETA for those is over the next few days.

 

A formal RCA will be released so you can reach out to your CSM/PSM from Cisco for more information on this by tagging the incident ID with them.

 

Thanks & Regards,

Arunabh.

 

ah right, yeah I am using anz1

thanks!

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: