Recovery alert not using correct values

We are using Telegram as a transport for LibreNMS alerts. When a recovery alert is issued by LibreNMS it uses the values from the most recent warning alert. This way the alert itself is a recovery but the values makes it seem there is still an ongoing issue.

For instance in this case an alert is issued for port congestion, the threshold for this alert is 85%. The first alert shows an inbound utilization of 94.63%, the recovery alert again shows this value while in fact it is < 85% (otherwise no recovery alert would’ve been send out).

Any input on this would be greatly appreciated!

libre

Anyone? Some input on this would be greatly appreciated