University of Oregon

Systems - Services - VMware - Standing SLA

Audience
Faculty/Staff
Researcher
Student
GTF

VMware Service

Standing Service Level Agreement­

Overview

This service level agreement (SLA) is between Information Services (IS), and any Administrative, Academic, or Research Unit that is part of the University of Oregon. Under this SLA, Information Services agrees to provide access to the services outlined in this document.

This SLA documents agreed-upon systems and services, covers performance and reliability targets objectives, outlines escalation processes, and serves as an invoice for financial transactions between parties.

Purpose

In support of the academic mission and administrative functions of the University of Oregon, Information Services provides the following servers which are to be used by the University of Oregon to provide services to the community at large.

Known and supported uses of these servers include:

  • Production Servers
  • Development Servers
  • Test Servers
  • Research Servers

Duration of Agreement

This standing agreement covers all services running on the Information Services VMware cluster. Any modification to this agreement will be posted to the Information Services website at the following URL:

Definition of Scope

The following servers and appliances are covered by this agreement for the following uses:

Hostname Description Specs Acquisition Date Approved Use
Acadesx-11 PowerEdge R710 Dual Intel X5660/144GB 9/2011 Production Cluster Member
Acadesx-12 PowerEdge R710 Dual Intel X5660/144GB 9/2011 Production Cluster Member
Acadesx-13 PowerEdge R720 Dual Intel E5-2670 0/144 GB 8/2013 Production Cluster Member
Acadesx-14 PowerEdge R720 Dual Intel E5-2670 0/144 GB 8/2013 Production Cluster Member
Acadesx-15 PowerEdge R720 Dual Intel E5-2650 v2/144GB 5/2014 Production Cluster Member
Acadesx-16 PowerEdge R720 Dual Intel E5-2650 v2/144GB 5/2014 Production Cluster Member
Acadesx-17 PowerEdge R730 Dual Intel E5-2699 v3/384GB 12/2015 Production Cluster Member
Acadesx-18 PowerEdge R730 Dual Intel E5-2699 v3/384GB 12/2015 Production Cluster Member
Acadesx-19 PowerEdge R730 Dual Intel E5-2699 v3/384GB 12/2015 Production Cluster Member
Acadesx-20 PowerEdge R730 Dual Intel E5-2699 v3/384GB 12/2015 Production Cluster Member
Acadesx-21 PowerEdge R730 Dual Intel E5-2699 v3/384GB 12/2015 Production Cluster Member
Acadesx-22 PowerEdge R730 Dual Intel E5-2699 v3/384GB 12/2015 Production Cluster Member

Additional Dependent Service Offering

IS UOCloud Storage Service - https://it.uoregon.edu/systems/services/storage

Please note that all equipment is dependent on fundamental infrastructure such as: redundant power feeds, battery backup systems, redundant cabling (copper and fiber-optics) infrastructure, redundant network connections to upstream network connectivity providers, and network switches not covered by this service level agreement. Information Services considers these circumstances prior to making claims about the availability of equipment covered by this agreement.

Supported Software

VMware vSphere Webclient

Duties:

Duties of Information Services

1. Responsible for hardware, firmware, and VMware software updates
2. Responsible for hardware troubleshooting, diagnosis, and resolution
3. Responsible for performance and availability monitoring including after-hours response to system outages related to facilities, hardware, or VMware malfunction
4. Responsible for documentation of configuration and maintenance procedures and processes concerning VMware environment
5. Responsible for delegating appropriate levels of access necessary for the purposes of Systems Administration for client

Duties of VMware client

1. Responsible for backup and restore functionality to ensure continuity of data and restoration of service in the case of a disaster that impacts VMware client.
2. Responsible for the installation of VMware tools and updates to it.
3. Responsible for the application of all patches and updates to the guest operating system
4. Responsible for the application of all patches and updates to any and all software installed on top of the guest operating system.
5. Responsible for performance and availability monitoring including after-hours response to system outages related to guest server configuration problems or disruptions
6. Responsible for documentation of configuration and maintenance procedures and processes concerning guest environment
7. Responsible for timely response to all security concerns
8. Renewal of all necessary support agreements covering installed software and/or operating systems licensing (Microsoft Windows and Redhat Linux base licensing is included as part of the service, CALs are not).
9. Responsible for proper alignment of all partitions configured in any virtual hard drives.
10. Responsible for supplying maintenace windows within 6 weeks of request for VM right-sizing operations.

VM Sizing

VMs running in the environment must be right-sized.  Resources must be assigned to meet the needs of the application (but not over).  This ensures that resources are used effectively.  Sizing for running machines will be based on observed running operation and will account for any required buffer for failover operations.

Right-sizing is compulsory for VMs provided from IS-funded resources.  When requested, a maintenance window (which may involve a reboot) must be provided to make any changes to the VMs.  That window must be within 6 weeks of the request.  Failure to provide a maintenance window for your specific VM will result in maintenace being performed in the next available regular service maintenance window.

Guest OS Support

Systems will deploy new VMs with a gold master template if desired.  Available templates are:

Windows 2008R2
Windows 2012R2
Windows 7
Windows 8.1

Redhat5 (32 or 64bit)
Redhat6 (32 or 64bit)

If you would like to run another OS, and its supported by our current version of VMware ESX, we can provision you a blank VM for you to install the OS onto.

Maintenance periods

Information Services – Systems maintenance will occur between 5am and 7am on Tuesday mornings. Management servers may be temporarily unavailable during this time period unless otherwise noted in this document.  Maintenance done outside this window will be announced via the mailinglist (see FAQ for details) at least 24 hours in advance.

Typical work performed during systems maintenance periods:

* Applying operating system patches
* Rebooting active production servers
* Rebooting fail-over servers
* Changing or modifying firewall rules
* Upgrading application specific software
* Updating hardware firmware levels
* Updating dependent software
* Tested and approved system configuration changes

Every effort is made to impact service availability of production instances as little as possible, however during the duration of regularly schedule maintenance periods, services running on hardware covered in this document may be unavailable.  Impacts to VM operation will be scheduled and announced in advanced to the mailing list.

Network Services reserves separate maintenance windows to perform regular network upgrades.

While the systems will still be functioning during these time-periods, users may not be able to access the servers during this time due to network interruptions.

Information Services reserves the right to perform maintenance during these periods without prior notification to impacted parties.

Incident Reporting and Escalation:

Tier 1:

Information Services – Systems and Operations

Tier 2:

Leo Clark - Storage and Virtualisation Lead

Tier 3:

Micah Sardell – Associate Director of Systems and Operations

Tier 4:

CIO

Escalation:

Initial point of contact for all system availability problems will be via the systems email address or via the systems telephone hotline.

Software Change Requests:

Initial point of contact for all systems configuration changes will be with a designated member of the Information Services – Systems staff assigned after initial contact via email or the Systems Hotline.

Information Services may refuse to implement a change if they determine that change broadens the scope of service, or if they determine that it adversely affects other aspects of service availability.

Service Expectations:

Application users can expect approximately 99.9% service availability (Approximately 8.7 hours of unplanned outage) measured over the time period of a year, excluding anticipated outages and downtime performed during maintenance periods.

Service availability will be measured using Nagios service monitoring and records will be retained for a year. We will monitor the VMware physical hosts as well as the availability of the Netapp filers.

Overall system performance statistics will be monitored and graphed for a period of one year through the VMware vSphere Client. Currently Systems uses vCenter to graph and archive system performance data.

Please note: This infrastructure is not specifically designed to meet PCI or HIPAA standards.  Consumers of this service are responsible for evaluation of the service for compliance with any regulations required by their data.  More information available on request.

Penalties:

No penalty for missing this availability expectation will be enforced at this time. This service expectation should be used for planning purposes only.

Fees:

Please see contact Systems and Operations to discuss an MOU with pricing details.

*Note that this does not reflect the actual cost of services rendered but rather a contribution towards payment of services.