API performance went slightly haywire this week in the Open Banking and the US Government categories. Buckle up, it’s gonna be a bumpy report!
Open Banking API Performance Issues
First up we have a fairly significant decrease in TSB’s API performance. It slid into the Red Zone with a CASC score of 5.95. This was owing to a large number of slow outliers affecting all endpoints through the week. This was most obvious during the weekday.
Next up we have Barclays Personal Sandbox in the Open Banking UK Sandboxes collection, which was in the Red Zone with a CASC score of 5.95. This was because of a number of clusters of slow outliers affecting some endpoints with other endpoints experiencing a high percentage of failed calls through connection errors with no HTTP status code being returned.
Now, sandbox API performance isn’t necessarily a big deal. But Barclays Personal in the Open Banking UK Production collection also decreased by 16% from 9.05 (in the Green Zone) to 7.58 (in the Yellow Zone).
This was owing to a cluster of slow outliers that affected most endpoints on 25 March. There was also a maintenance outage on 26 March.
Rounding out API performance issues in the Open Banking UK Production collection is New Day, with two separate situations.
NewDay – Amazon Mastercard in the Open Banking UK Production collection decreased by 20% from 8.73 (in the Green Zone) to 7.00 (in the Yellow Zone). This was owing to a cluster of slow outliers affecting all endpoints at midnight UTC each day, presumably linked to a system reset of some kind.
NewDay – Aqua in the Open Banking UK Production collection decreased by 22% from 8.74 (in the Green Zone) to 6.80 (in the Yellow Zone). This was owing to a cluster of slow outliers affecting all endpoints at midnight UTC each day, presumably linked to a system reset of some kind.
Now, let’s move on to our facorite repeat offender, the US Government.
US Government API Performance Issues
As we have mentioned before, the US Government needs some serious attention paid to their APIs. This week, we start with the FAA.
The FAA in the US Government collection was in the Red Zone with a CASC score of 4.50. This was owing to a cluster of failures and slow outliers on 25-26 March 2023 affecting all endpoints. Failed calls returned with HTTP status code 500 Internal Server Error responses and the error message “Error in Assertion Processing.”
Dept of Energy in the US Government collection was in the Red Zone with a CASC score of 4.05. This was owing to a cluster of slow outliers beginning on 24 March 2023 with some failed calls with the HTTP status code 504 Gateway Timeout being observed.
Veterans Affairs in the US Government collection was in the Red Zone with a CASC score of 3.71, the worst score of the week. This was owing to a continuous outage beginning on 25 March 2023 at 4:52 UTC and continuing to 27 March with HTTP status code 503 Service Unavailable being returned and the message “We can’t find the page you requested, but we’ve been notified about the problem.” (Note the endpoint started working again on 27 March.)
Bureau of Labor Statistics in the US Government collection decreased by 13% from 9.20 (in the Green Zone) to 7.98 (in the Yellow Zone).
This was owing to intermittent failed calls with no HTTP status code returned and a cluster of such calls on 25 March that failed much more quickly.
Bringing up the rear is the Dept of Justice in the US Government collection, which decreased by 23% from 8.72 (in the Green Zone) to 6.68 (in the Yellow Zone).
This was owing to many slow outliers during the week and a cluster of failed connections on 25 March with no HTTP status code returned and the system error message “A problem while resolving domain name” generated.
