Ops Log tag
#Monitoring
Reviewed Ops Log notes tagged Monitoring, written by Michal Jatczak with source links, test context and operational checks.
4 reviewed notes
A daily offboarding check can prove closure only as a bound
A leaver check that runs once a day can report closure only as a bound between two runs. A Graph 429, including one inside a batch that returns 200, and a membership read that returns nulls can each make a degraded run look clean, so an absence counts only from a run that read cleanly.
Azure Cost Management can return 429 with 597 of 600 tenant queries unused: the exhausted bucket is the client type
A Cost Management query returned 429 on 2026-08-14 while the response header for the documented per tenant quota still reported 597 of 600 hourly queries left. The exhausted bucket belonged to the caller's client type, and the same request body with a ClientType header returned 200 seconds later.
Windows 11 24H2 stops receiving updates a year earlier on Pro than on Enterprise, from the identical build
Windows 11 version 24H2 ends updates on 2026-10-13 for the Pro family and on 2027-10-12 for the Enterprise family, from one table row and one identical build. Count the fleet by edition, not by build, before the earlier date arrives.
Running Prometheus, Loki and Alertmanager on one VM: configuration, retention and cost
A single-VM Prometheus, Loki, Alertmanager and Grafana deployment that starts on the first attempt: the retention flags that belong on the command line, the disk-sizing formula that decides the machine class, alert rules with the false-page guards in place, and a cost model built from published list prices instead of asserted savings.