Network Service Agreement

  • If network administration determines that the network (or a specific file, print, or application server) must be shut down during business hours, NetHelp support will notify the DSAs to instruct users to save all work (if possible) and log off the network. In the event that the DSA is not available, the Data Trustee will be notified. Every effort will be made to minimize the interruption.
  • In the event of a network server crash, there is usually no advance notice. When service is restored, data is available only from the point of the last save.
  • Once an incident has been resolved and network services are once again fully functional, DSAs or Data Trustees will be notified to inform users that they may log back in.
  • Where feasible, the ESO network infrastructure had been equipped with various levels of fault-tolerance and redundancy. In the event of a system failure, this would allow for the continuation of some services without a significant performance impact or interruption to users.
  • If a system failure is localized only to our network and depending on the severity of the failure, it may take up to 4 hours before services are fully restored to normal operational levels. However, if the impact also involves the larger campus-wide network, it could take longer to restore services.
  • The network is equipped with Uninterruptible Power Supply (UPS) units which provide it with battery power for a limited time. These automatically engage when there is a power surge or outage that affects the building where the servers are managed. Specialized software then notifies network personnel of the situation and shuts down each server before battery power is exhausted. All servers will be brought back online as soon as power to the building is fully restored.
  • All maintenance, hardware upgrades, and software updates are performed on the primary servers during maintenance hours from 8 p.m. Friday until 4 a.m. Saturday.
  • The ESO network infrastructure is actively monitored for reliability, performance, and availability to reduce the possibility of system failure and minimize downtime.