add flaky_test decorator to rerun test_ceph_osd_slow_ops_alert

[DanielOsypenko]

1. try to rerun the queries or not to fail when Prometheus did not resp. It is a known issue that for couple minutes it may become unresponsive.
   E ocs_ci.ocs.exceptions.TimeoutExpiredError: Timed out after 300s running get("url"="[https://prometheus-k8s-openshift-monitoring.apps.j-011vup1cs33-t3.qe.rh-ocs.com/api/v1/alerts"](https://prometheus-k8s-openshift-monitoring.apps.j-011vup1cs33-t3.qe.rh-ocs.com/api/v1/alerts%22), "headers"={'Authorization': 'Bearer sha256~wthhAPuEnlh8YAHGbIAEqApRbE-af_I15UW_BuOBvTw'}, "verify"=False, "params"={'silenced': False, 'inhibited': False})
   https://reportportal-ocs4.apps.ocp-c1.prod.psi.redhat.com/ui/#ocs/launches/632/19989/968190/968203/log

2. Failed: failed to get 'CephOSDSlowOps' while workload filled up the storage to 0.85 percents
   https://reportportal-ocs4.apps.ocp-c1.prod.psi.redhat.com/ui/#ocs/launches/632/19958/967097/967110/log
   In this case it worth to rerun the test or wait more time to get the alert.

---

try to get this Alert on Compact mode, manually.

Add flaky decorator, same as - #9532