SLA for Deployment Cluster

Welcome to the Platform Compute Services Virtual Machine Development Cluster. This document serves as the guidelines and regulations for participation with our “deployment” service.

Guidelines

  • Virtual Machines are subject to being powered off or become unresponsive without notice.
  • Virtual Machine Snapshots are available upon request and require approval.
  • Virtual Machine Operating Systems may be installed via mounted ISO, Web Kickstart/Foreman or WDS/SCCM.
  • Windows based Virtual Machines must have Crowdstrike Antivirus installed.  OIT has obtained enough licenses for all Windows-based guests hosted on OIT’s VMware farms.
  • Power off virtual machine when not in use
  • Destroy or delete virtual machine if determined it is not needed or project retires.

Procedure for Migrating Virtual Machines to Production Cluster

  • Power down the virtual machine.
  • Send a request to oit_platform@help.ncsu.edu.
  • Virtual Machine settings will be reviewed.
  • If Virtual Machine settings are correct, the system will be migrated.
  • Please be sure Virtual Machines are configured as follows:
    • Virtual Machine named with Fully Qualified Domain Name.
    • VMware Tools is installed.
    • VMware Tools settings:
      • Time sync checked on Windows and not checked on Linux.
      • install VMCI Driver.
    • Appropriate paravirtualized adapters are installed (at least VMXNET3 for all OSes).
    • <4 vCPUs (unless 4+ processor approval has been granted).
    • Hot add/plug settings enabled for CPU and Memory.

Exception Policy

The Deployment Cluster is not intended for long-term use. Virtual Machines must be transitioned out of the cluster within 90-days of creation unless written approval of extension is granted by the Platform Compute Services team.

Power off Policy

Platform Compute Services reserves the right to power off virtual machines if additional resources are needed to preserve the integrity of the cluster.

Deployment Cluster Availability

  • The deployment cluster is not available to all customers, and it is intended to be used for short periods of configuration and tuning, it is not for production use.  All production Virtual Machines should be migrated to the Production Cluster, which is monitored for 24×7 service.
  • The deployment cluster is not monitored outside normal business hours (8am-5pm EST Monday-Friday), and on-call staff are not expected to resolve issues in deployment off-hours.
  • Planned outages are announced via Sysnews for the production cluster, but notice may not be available for the deployment cluster.
  • Though we strive to minimize disruptions to the Deployment Cluster, it is subject to unscheduled reboots and outages at any time.

Avamar Backups

Virtual Machines running in the deployment cluster should not be configured for Avamar backups.  Backups discovered to be originating from Virtual Machines operating in the Deployment Cluster will be deleted and will require reactivation upon moving back to the Production Cluster. System Owners will be notified upon deletion of Avamar Backups.