Motivation for Downtime Policy
CSEHelp is charged with maintaining a large heterogeneous networking of computing and storage, all of which is directly connected to the Internet. Software and Operating systems used on these systems are constantly receiving updates (patching) to fix functionality and more importantly security. Timeliness of distribution of these updates is crucial in keeping would-be intruders out of CSE systems. Software updates can also be critical to maintaining stability of the overall system. This policy provides the mechanisms necessary for CSEHelp to consistently maintain the stability and integrity of CSE computing resources.
Downtime Policy
We classify downtimes in three (3) categories:
- Regular Maintenance Window
- Emergency
- Special
In general downtimes will be announced at least 24 hours in advance. Emergency downtimes (especially in cases of hardware failure or security breach) will be announced when they occur.
Notices will include:
- Date/Time
- Systems/Users affected
- Expected duration
- Purpose
- Description of any user required preparation
- Description of any user required recovery
Regular Maintenance Window
The purpose of the regular maintenance window is to establish pre-scheduled downtimes in order to maintain CSEHelp hardware and software systems in a timely manner.
On the second Thursday of each month CSEHelp reserves 7-10pm for downtime associated with regular maintenance of systems (including software upgrades, security patching, etc.)
For each maintenance window there will be one established alternate time. For
downtimes related to individuals desktop systems for whom the regular time would
be disruptive to a time-sensitive project, the window can be delayed to the
following Tuesday 7-10pm. These individuals need to contact CSEHelp before 5pm on the day of the primary window to request the alternate time.
If the need for a large or complicated upgrade arises. the 7-10pm window may be extended by CSEHelp. Any extensions will be noted in the downtime announcement.
Emergency
Emergency downtime can occur due to severe hardware/software failure or security breach. These downtimes are unavoidable and induced by outside forces. Announcements for emergency downtimes will be made as quickly as possible after the failure or breach has occurred. In cases where the downtime affects email service a note will be posted on the CSEHelp status page on the CSE departmental web server.
Special
Special downtimes will be scheduled for large projects that do not fit in the regular maintenance window. These downtimes will be announced well in advance so that all users can plan on the outage.
 |