Updated: July 21, 2008
The Office of the Vice President for Information Technology (OVPIT), primarily through University Information Technology Services (UITS), provides myriad “common good” services in support of institutional operations and individual users on all campuses. In support of detailed ongoing quality and stewardship processes (primarily “Activity Based Costing,” or ABC), nearly 180 of those services have been defined, with service unit and service manager annotated.
Service units must ensure that there are appropriately detailed recovery plans for each service, should the hosting data center be lost. The ABC index is a reasonable mechanism to ensure that all units have developed and that they maintain these plans for services assigned to them – “Service Recovery Plans.” Collectively, these Service Recovery Plans are a main component of the overall OVPIT/UITS Disaster Recovery (DR) Plan.
This document outlines the processes for the creation and maintenance of Service Recovery Plans. This process is comprised of:
· Annual Service Recovery Plan review process.
· Mechanisms for ensuring Service Recovery Plans are updated as services change.
· Mechanism for collecting Service Recovery Plans for new services.
· Mechanism for removing Service Recovery Plans for discontinued services.
Service Recovery Plans for OVPIT/UITS services are collected via a standard survey/template sent to each service manager. The information requested in the template includes:
· Technical architecture,
· Inventory of interfaces to IU internal services,
· Interfaces to organizations external to IU,
· Recovery strategy, and
· Detailed steps required in support of the recovery strategy.
Completed Service Recovery Plans can vary in content depending on the nature of the service. For example, a recovery plan for restoring services related to consulting, which is a more human-centric service, would reflect entirely different procedures than a plan covering the recovery of the wireless network, which is a hardware-centric service; the steps contained in a plan to recover an application that utilizes redundant hardware in an alternate location would reflect different process steps than an application that would be recovered on new equipment obtained from a vendor.
OVPIT/UITS Disaster Recovery plan documents, including Service Recovery Plans, are stored on an OVPIT server housed in the data center in the IU-Bloomington Wrubel Computing Center (WCC), and backed up via current standard methods (e.g., Tivoli Storage Manager). Copies of the backups are stored in the WCC data center and in the data center located in the Informatics and Communications Technology Complex (ICTC) at IUPUI. In the event that the OVPIT server must be restored, these documents would be restored along with other materials.
Additionally, electronic copies of the documents are stored on CDs that are housed in the Franklin hall room 300 disaster recovery storage cabinets, ICTC disaster recovery storage cabinets, and Bloomington machine room vault.
The steps outlined below represent the process for annual review of OVPIT/UITS Service Recovery Plans. This review will take place annually.
1) The OVPIT disaster recovery coordinator will develop and maintain an inventory of current plans. This inventory list will be stored on the OVPIT server with the master copies of OVPIT/UITS plans.
2) The OVPIT disaster recovery coordinator will obtain the current ABC service inventory from the OVPIT/UITS Finance Office.
3) Once both lists are obtained, they will be compared in order to identify:
a. New services added during the most recent ABC review process,
b. Changes in service “ownership,” and
c. Changes to individuals responsible for service.
4) The list resulting from the comparison becomes the population for the review of Service Recovery Plans; the OVPIT disaster recovery coordinator will then create whatever instruments (documents, tracking lists, etc.) as might be necessary to guide the review.
5) Current Service Recovery Plans will be copied to a location to which service unit managers have ready access, and organized to facilitate easy review.
6) The OVPIT disaster recovery coordinator will request that each service manager review their plans, provide guidance as to what the plans must contain, and set a date by which that review is to be completed.
7) Upon receipt of updated plans, the OVPIT disaster recovery coordinator will review them to ensure that they are complete. If not, the plans will be sent back to the service manager with questions and concerns annotated. There may be multiple iterations until the submitted plan is adequate.
8) Once the final version of the plan has been received, the document will be appropriately filed.
9) Previous plans will be retained in archive folders on the OVPIT server for 3 years.
10) Once all plans have been collected, copies of all plans will be written to CDs. Copies of the CDs must then be stored in:
a. Franklin hall room 300 disaster recovery storage cabinets,
b. ICTC disaster recovery storage cabinets,
c. WCC data center vault.
11) CDs containing plans from the prior year should be removed from these locations and destroyed.
12) Copies of plans placed in locations to facilitate service manager review should be deleted.
Services change throughout the year. Service unit managers must consider how these changes affect the associated Service Recovery Plans. The standard change management process will be leveraged in order to ensure that this is done.
Change request forms will reflect an opportunity for the service unit to indicate that they have appropriately addressed this aspect of implementing the change, by asking (in this or similar language):
“Does the change represented on this form impact the contents of your Service Recovery Plan for this service? If the answer is Yes, have you coordinated the change in the Service Recovery Plan with the OVPIT disaster recovery coordinator?”
“If this change represents a new service, has a Service Recovery Plan been finalized with the OVPIT disaster recovery coordinator?”
If these questions are not answered to the satisfaction of the change board, the request should be denied (with appropriate consideration to other factors).
A similar process must occur in other forums in which changes are discussed and where changes could affect Service Recovery Plans, but those changes might not be represented in the broader change management process. (E.g., Enterprise Systems Integration Team meetings.)
If the service is new, the following steps should occur:
1) The service owner should contact the OVPIT disaster recovery coordinator to obtain a template that outlines appropriate plans components and instructions.
2) The OVPIT disaster recovery coordinator and the service unit manager will interact to ensure that the complete and appropriate information is provided in the plan.
3) Once the Service Recovery Plan is complete, the OVPIT disaster recovery coordinator will:
a. Add service information to the DR master inventory list. If the service is not reflected on the current ABC services list, the inventory list should reflect that.
b. File the Service Recovery Plan in the OVPIT document repository.
c. Update and distribute new CDs with updated DR documents:
i. Franklin hall room 300 disaster recovery storage cabinets,
ii. ICTC disaster recovery storage cabinets,
iii. WCC data center vault.
If an existing service is being changed and those changes may affect the Service Recovery Plan:
1) The service unit manager must contact the OVPIT disaster recovery coordinator to obtain a copy of the plan currently on file, and discuss the change and impact to the current plan.
2) OVPIT disaster recovery coordinator and the service unit manager will then interact to ensure that the plan reflects the necessary adjustments.
3) Once the Service Recovery Plan is complete, the OVPIT disaster recovery coordinator will:
a. Add service information to the DR master inventory list. If the service is not reflected on the current ABC services list, that fact should be annotated on the inventory list.
b. File the Service Recovery Plan in the OVPIT document repository.
c. Update and distribute new CDs with updated DR documents:
i. Franklin hall room 300 disaster recovery storage cabinets,
ii. ICTC disaster recovery storage cabinets,
iii. WCC data center vault.
The services list associated with the ABC process is maintained by the OVPIT Finance Office. The individual assigned to manage that process will coordinate with the OVPIT disaster recovery coordinator to ensure that changes to that list are shared, so that the above appropriate process is invoked as necessary, where changes made have bypassed standard change management procedures.
If a service is removed from the ABC services index, the OVPIT disaster recovery coordinator will confirm with the service unit manager that the service is, indeed, no longer active, and if so will archive the associated Service Recovery Plan(s) on the OVPIT server, and remove that plan from the distributed CDs at the next reasonable opportunity.