multiple resolved alerts

What did you do?
Exec flow script to send alerts to alertmanager every 1m

alerts='[
  {
	"labels": {
		"instance": "10.70.58.29:9100",
    "alertname": "DiskRunningFull",
		"instance_name": "host1"
	},
	"annotations": {
		"trigger_value": "0"
	},
	"startsAt": "2021-10-08T07:13:40.342408949Z",
	"endsAt": "2021-10-09T07:14:40.342408949Z"
},
{
	"labels": {
		"alertname": "DiskRunningFull",
    "instance": "10.70.58.30:9100",
		"instance_name": "host2"
	},
	"annotations": {
		"trigger_value": "0"
	},
	"startsAt": "2021-10-08T07:13:40.342408949Z",
	"endsAt": "2021-10-08T07:14:40.342408949Z"
}]'
curl -XPOST -d"$alerts" http://localhost:9093/api/v1/alerts

What did you expect to see?
Only one resolved alert for host2.

What did you see instead? Under which circumstances?
Received multiple resolved alerts for host2 every 1m

Environment

• System information:

Linux 4.15.0-142-generic x86_64

• Alertmanager version:

0.19.0 & 0.23.0

• Alertmanager configuration file:

global:
  resolve_timeout: 2m

route:
  group_by: ['alertname']
  group_wait: 10s
  group_interval: 10s
  repeat_interval: 1m
  receiver: 'web.hook'
receivers:
- name: 'web.hook'
  webhook_configs:
  - url: 'http://127.0.0.1:5001/'
inhibit_rules:
  - source_match:
      severity: 'critical'
    target_match:
      severity: 'warning'
    equal: ['alertname', 'dev', 'instance']```

* Prometheus configuration file:

insert configuration here (if relevant to the issue)

* Logs:
no logs output

Start by reproducing the behavior with the provided curl request to the Alertmanager /api/v1/alerts endpoint and the shown route configuration. Trace how repeated resolved alerts are grouped and repeated, then verify that the host2 scenario produces only one resolved alert while preserving expected behavior for the other alert.