ITS Data Center Infrastructure Dates

Data Center infrastructure work can require extended outages for all services in the data center. Examples of such work include changes to our electrical, mechanical, network, or firewall infrastructure. The most visible example of such work was the electrical work in June 2007 which required that we shut down the entire data center for several hours.

Scheduling infrastructure work requires considerable coordination.  It is important that we schedule the work when it is least intrusive on the entire user community. It is important that we have key staff and vendors on site.

While we remain in the current data center, it is anticipated that we will need to do infrastructure repairs, upgrades and preventive maintenance that will require an outage. Therefore, we would like to establish some windows that we could use for planning purposes.

Identifying maintenance windows is difficult with our complex environment, varied stakeholders (academic, administrative, research, collegiate, and departmental) and their diverse business cycles. Our goal is to identify three windows during the year where we can schedule extended outages. Ideally these windows would be identified well in advance, hopefully as part of our annual Campus IT planning cycle. We would only use the maintenance windows if necessary, and we would provide advance notification of our intent to use the maintenance window. As always we would minimize actual downtime.

After conversations with administrative and academic units, we are proposing the second Saturday in January, June, & October for 2010.

• Saturday, January 9, 2010
• Saturday, June 12, 2010
• Saturday, October 9, 2010

Work Completed on the 2009 Data Center Maintenance Dates

• Saturday October 10, 2009
  • Maintenance date not needed.
• Saturday, June 13, 2009, 5AM - 6:30AM.
  • Supervisor cards were replaced in two networking switches in the primary data center.
  • Software and O/S updates were applied to the firewall.
• Saturday, January 10, 2009, 6AM-9AM
  • Software updates and configuration changes were made to the network routing infrastructure in the primary data center. These changes were necessary in order to improve the reliability of network connections, both within the data center and to the rest of the campus network.

Work Completed on the 2008 Data Center Maintenance Dates

• Saturday, February 16, 2008
  • Primary data center operated via UPS & portable generator for electrical work to enable the installation of a permanent generator.
  • Most test and development servers were shutdown to reduce the load on the portable generator.
• Saturday, June 14, 2008
  • Due to the Flood of 2008, we cancelled the work planned in our primary data center to replace an electrical manual transfer switch with an automatic transfer switch in preparation for a permanent generator.
  • On Sunday, June 15, 2008, we lost power to our primary data center for several hours requiring a portable generator be placed into operation. We have scripted our server shutdowns in order to quickly shed power during such events. We have five levels and we were at level three when the portable generator came online. (Level one includes test servers. Level two includes development servers. Level three includes some production servers.)
  • On July 31, 2008 we completed the installation of a permanent generator. This installation is particularly important as electrical power is currently provided to our primary data center via a radial feed that is dependent on MidAmerican Energy.

Work Completed on the 2007 Data Center Maintenance Dates

• Saturday, June 9, 2007, 6AM - Noon
  • Replaced the electrical infrastructure in the primary data center by upgrading circuit breakers, increasing wiring capacity, and upgrading the capacity on the Uninterruptable Power Supply. This work required a six-hour outage for all central computing services provided by the primary data center.

 

Return to ITS Alerts