Cirrus Hosting (WHG) automated updates will crash your server – Zero monitoring and corporate evasion
I have been a loyal customer of Cirrus Hosting since 2010 (16 years of billing history), paying for a Managed VPS service. Recently, my production environment suffered a catastrophic 24-hour database downtime entirely due to infrastructure failure and faulty automation on Cirrus’s side.
On July 6th, their automated systems executed a global update, upgrading the node to cPanel version 134.0.43. During this process, their internal migration scripts malfunctioned, overwriting my stable configurations and forcing an amateur, destructive default blueprint directly into my account files: memory_limit = 2G and max_execution_time = 600s.
Setting a 2GB limit per script on a Managed VPS node with only 4.7GB of total RAM is an amateur infrastructure configuration error. Predictably, this triggered continuous Out-Of-Memory (OOM) kernel events, completely crashing the core MariaDB service.
Despite paying for a premium "Managed" service, Cirrus's monitoring infrastructure completely missed this 24-hour business outage. Swap memory remained 100% exhausted. As a customer with no root access, I had to personally use the terminal/CLI to isolate the root cause, clear the swap, and literally dictate the safe production values (512M / 300s) to their Level 3 technical support to stabilize the node.
The Retention department Specialist explicitly promised an account review for appropriate compensation options (a 3-month service credit) once stability was achieved. However, the Customer Advocacy Team Lead abruptly denied it, using cPanel UI anomalies to falsely claim that the 2G limit was "client responsibility," dismissively stating that because my account is canceling on August 2nd, no resolution is appropriate.
The internal server logs indisputably record that their automated system update forced this destructive configuration. This is the classic WHG buyout pattern: acquiring functional hosts, cutting support quality, breaking servers with master templates, and then blaming the user. Avoid Cirrus Hosting and World Host Group entirely.
Internal support tickets and system logs documenting this outage have been fully preserved and provided to management.
[UPDATE July 14]: Corporate silence continues. My internal support tickets regarding the catastrophic 24-hour database failure have been completely ignored and kept in 'In Progress' status since Friday. Cirrus Hosting is clearly stalling to avoid contractually promised compensation before my account cancels on August 2nd. I have formally demanded the senior management to pull the node’s update logs (/var/log/cpanel/updatelogs/) from July 6th, which indisputably record their automated script failure. If this corporate evasion and total lack of managed support persists, this case and full log history will be escalated directly to the CCTS and BBB for breach of SLA.
6 de julho de 2026
Avaliação não solicitada