A Major Midwestern Academic Medical Center Replaces Collapsing Firmware Server with Endpoint Manager R6

A major Midwestern academic medical center: a public university health-sciences campus paired with a large clinical enterprise. The combined organization is a tertiary and quaternary referral destination — Level I trauma, solid-organ and bone-marrow transplant, comprehensive cancer care, infectious-disease and biocontainment capability, cardiovascular and neurosciences programs alongside colleges that train physicians, nurses, pharmacists, and allied-health professionals. 

A system of this class typically operates two hospitals totaling about 800 licensed beds, 70 or more specialty and primary-care clinics, roughly 10,000 health-system employees, more than 1,000 affiliated physicians, and several thousand university faculty, staff, and learners on a shared campus of hospitals, research towers, and outpatient centers. Annual activity is on the order of a million patient encounters, including high transfer volume from community hospitals across the region. 

Voice is not an office amenity here. It is how trauma bays, ICUs, transplant coordinators, clinics, and a statewide training network reach one another. When firmware and settings distribution fails during a reboot wave, clinical communications stall on a campus that cannot wait for a truck roll. 

  • Challenge

    The production problem was specific and visible. The medical center was using a general-purpose Apache web server to host Avaya firmware and 46xxsettings files. Any widespread reboot of IP phones after a power event, a network change, or a planned refresh created a stampede of devices requesting files at once. The web server could not absorb that burst. Phones recovered slowly or not at all, and the telecom team had no purpose-built control plane for inventory, grouping, remote reboot, or policy-based firmware. 

    At the same time the estate was large and mixed: 

    • More than 10,000 Avaya endpoints across hospitals, clinics, academic buildings, and research facilities.
    • H.323 devices still in production, requiring Avaya Application Enablement Services with SMS links to Communication Manager.
    • Multiple Avaya Aura serving sites (Communication Manager, System Manager, Session Manager, AES) that had to be represented as first-class objects, not spreadsheet tabs.
    • A security-conscious academic IT organization that required enterprise endpoint-security agents (Tanium and FireEye) on Linux management hosts —without testing those agents first on the production pair. 
    • A lean operating model: implementation had to be remote, using desktop sharing with the customer’s on-site team. The supplier did not requireVPN access into the medical center’s network.

    A generic web server is not an endpoint-management platform. The medical center needed a dual-server, high-availability appliance sized for the full production load, plus a safe lab path for security tooling.

  • Solutions

    INEMSOFT Endpoint Manager R6.x was deployed as an on-premises dual-server high-availability (HADR) pair on customer-supplied VMware virtual machines. The production design was configured to support 10,100 devices. A separate lab server was recommended so Tanium and FireEye could be downloaded, deployed, and tested against Endpoint Manager on CentOS R8 or later — creating a buffer between security-agent QA and the live clinical pair. 

    Standard implementation covered a joint project plan, dependencies, the IPT workbook for the customer’s Avaya Aura environments, a second-day operating model for firmware and settings, remote install and configuration of the two production OVA instances, and acceptance testing with the customer’s resources. Day-to-day operations were then transitioned to the customer support team. 

    Avaya AES with SMS links remained a customer prerequisite for H.323. The customer prepared the bulk-load CSV of devices and groups (names, locations, numbers, and extensions per group) in the supplier format.

    Key Results

    • Apache bottleneck removed Firmware and settings are vended by a purpose-built HA pair instead of a general-purpose web server that collapsed under reboot storms. 
    • Production scale with HA Dual-server HADR configured for 10,100 devices across hospitals, clinics, and academic buildings. 
    • Remote lifecycle control Telecom can vend firmware and settings and restart devices without walking the campus or driving to satellite clinics. 
    • Security-safe management plane Tanium and FireEye can be qualified on a lab server before they ride on the production Endpoint Manager hosts.
    • Repeatable onboarding Maps and profiles aligned 46xxsettings to model, subnet, group, location, and business unit so new sets follow policy instead of tribal knowledge.
    • Closed-campus delivery Implementation completed remotely via desktop sharing. No supplier VPN into the medical-center network.

Other Case Studies of interest

Schedule a demo today

Learn more about our solutions with a demo.

schedule a demo
How can we help you?

Connect with an iNEMSOFT expert to get started.