TouchWorks Failure Points

Server Outage

• Any critical point of failure goes down, the application may be unavailable or functionality may be limited

DB Server

Clustered Environment

• If SAN connection fails or SAN becomes unavailable, DB would be offline and users would be unable to use application
• If both nodes fail, DB would be offline and users would be unable to use application
• If one node fails, users would lose connection while the passive node is taking over it's duties. This typically takes 30-90 seconds according to Microsoft. TouchWorks VOE service on Web servers should reestablish their connection to the DB, but may need to manually restart that service if the services don't establish the connection automatically.

Non-Clustered

• If any component of the server failed, such as the NIC cards, Hard drives, processors, etc, the DB would be offline and users would be unable to use application

Web Server

Load Balanced Environment

• If Load Balancer device fails, all users will lose connection to the DB server. Could allow them to hit web servers directly if the downtime is expected to be an extended period of time.
• If one web server fails, users connected to the server will lose their connection, but they should be able to reestablish their connection to the properly functioning web server immediately. There may be some loss of data. You also may add too much strain to one server if you do not have enough servers in the farm to handle the number of users with a server down.

Non-Load Balanced Environment

• If the web server fails, all users will lose connection to the applicatin. Users should be instructed to follow downtime procedures until the system is back online.

Scan Server

Clustered Environment

• If SAN fails the scan application will become unavailable and users will not be able to view scanned images within TouchWorks. All other functions would be ok.
• If both nodes fail, users would not have access to the Scan application and users would not be able to view scanned images. All other functions would be ok.
• If one node fails, users will be down while the passive node takes over. This should last 30-90 seconds according to Microsoft. During this time, scanned images will not be viewable in TouchWorks either. All other functions would be uninterrupted.

Non-Clustered Environment

• If any component of the server fails, the scan application would be unavailable and users would not have the ability to view scanned images in TouchWorks. All other functionality would be fine.

Interface Server

Clustered Environment

• If SAN fails, or both nodes of the cluster fail, the interfaces would be down. This would mean that Reg/Sched updates and any inbound or outbound interface would be down. Users would most likely notice arrived appointments or registration updates are not flowing to TouchWorks first. The users would still be able to complete their work on Pending appointments, but walk-ins would be difficult to tie to the correct encounter. The work around would be to create a new encounter to document the visit and correct the visit information once the interfaces were working again. Depending on the downtime, there could be a delay afterwards if there are a large number of transaction queued up.
• If both nodes fail, the same thing would occur
• If one node failed, the interfaces would be down for 30-90 seconds while the passive node takes over. The ConnectR service should start automatically, but any interfaces not set to auto-start would need to be manually started. The users would probably not notice this downtime at all.

Non-Clustered

• If any component of the server failed, such as the NIC cards, Hard drives, processors, etc, the Interface server would be offline and no data would be flowing into or out of TouchWorks.

Print Server

• If the print server goes down, Printing, faxing, SureScripts and the Delta process would be down. Print, Fax, and SureScripts jobs would queue up in the Database, but would not print until the Print server was back online. Users could print through the print preview function to their local printer if they needed to print during the downtime. this serves as a great backup. If patients need prescriptions soon and the print server is scheduled to be down for a while, users should be instructed to phone in prescriptions

Practice Management

• If the practice management system fails, users will be unable to arrive appointments. Any previously scheduled patient would be on the TW schedule and their visit could be documented normally. Any walk-in would have to have a new encounter created and once the system is back online, the information should be moved to the correct encounter (especially the note and charge)

Network

• If any portion of the network fails, anyone that is dependant on that link will not be able to access the system. The users will need to follow downtime protocol until the network is repaired.
• If the network fails within the datacenter users will experience downtime or limited functionality based on the components effected.

Power

• If power is lost at a site, users will be down unless UPS devices have been installed to ensure that critical components are protected.
• If power is lost at the datacenter, all users will be down unless UPS devices are in place to protect against this failure.