SLA Credit Request - Unplanned Host Failure Causing Data Loss on my server

Question

SLA Credit Request - Unplanned Host Failure Causing Data Loss on my server

Al Mokhlif Oud 0

Hello Azure Billing Team,

I am requesting SLA credits for an unplanned platform failure that caused extended downtime and permanent data loss on my production virtual machine.

Incident Summary:

On Wednesday, February 18, 2026, Azure detected that the physical host running my VM was degraded. The host failed catastrophically within hours, causing complete disk corruption and making my production server unrecoverable.

Downtime Details:

Start: February 18, 2026 at 1:21 PM UTC (4:21 PM GMT+3) — "Degraded performance detected (Unplanned)"
First crash: February 18, 2026 at 4:05 PM UTC (7:05 PM GMT+3) — "Unavailable (Unplanned)"
Second crash: February 18, 2026 at 7:20 PM UTC (10:20 PM GMT+3) — "Unavailable (Unplanned)"
End: February 19, 2026 — Health Event Resolved after backup restore
Total unplanned downtime: 8+ hours

Impact:

Complete filesystem corruption caused by host I/O failures
All processes failed with "Input/output error" — OS unable to execute any binaries
VM was unrecoverable — required full restore from backup dated 2/15/2026
3 days of permanent production data loss (customer data, database changes, application updates).

This incident was entirely caused by Azure's physical host failure. This has been confirmed by:

The VM's Resource Health history showing "Degraded performance detected (Unplanned)" and two "Unavailable (Unplanned)" events — all platform-initiated
A call with Microsoft support engineer on February 24, 25 2026, who confirmed the issue was on Azure's side and that the engineering team is scheduling a VM migration for March 20, 2026
A confirmation email from Microsoft acknowledging the platform failure (attached)

The Azure notification at 1:21 PM UTC stated the host was "potentially degraded" and gave a 30-day window to redeploy (by March 20, 2026). However, the host failed catastrophically just 3 hours later — well before any reasonable customer action could be taken. Azure's own recommendation presented waiting as a valid option.

I am requesting:

SLA credits for the 8+ hours of unplanned downtime in accordance with the Azure Virtual Machines SLA
Additional goodwill compensation for the permanent data loss, as standard SLA credits do not reflect the true severity of an incident where Azure's hardware failure destroyed a production filesystem and caused irrecoverable data loss

All health events are documented in the VM's Resource Health history. Please let me know if you require any additional information.

Thank you,

Al Mokhlif Oud

Himanshu Shekhar 4,025 Reputation points Microsoft External Staff Moderator

2026-02-25T20:33:06.35+00:00

Al Mokhlif Oud

Can you confirm whether the virtual machine was deployed as a single instance (not part of an Availability Set or Availability Zone) at the time of the incident, for Azure VM SLA validation?
Al Mokhlif Oud 0 Reputation points

2026-02-26T10:34:44.0433333+00:00

Hello,

The virtual machine wwwServer is deployed in Availability Zone 1 in the UAE North region. It is not part of an Availability Set.

The applicable SLA for VMs deployed in Availability Zones guarantees at least 99.99% uptime. The VM experienced 8+ hours of unplanned downtime on February 18, 2026 due to a platform-initiated host failure, which was confirmed by Microsoft support call.

Please let me know the next steps.

Thank you