Centralizing engineers can reduce the need for technical staff at every location, but it does not eliminate the need for someone to reach equipment, unlock rooms, or carry out a safe physical action. In Serguey Shinder’s first-person account of an organization with 39 sites and engineers based at two, a network fault took minutes to diagnose but hours to resolve because the site was not ready for a remote fix.
What happened at the 39-site organization
In his DEV Community article, Serguey Shinder describes an estate of 39 locations: 22 depots, two laboratories, 12 small sales offices, and one head office. Technical staff were based at two locations. Shinder says the organization made that choice in 2019 to save money, but had not replaced the physical response capacity once provided by regional technical posts. The article’s publication year is not established, and its account is not an independently verified audit.
The operational gap became clear when a switch failed at the largest depot and stopped dispatch. According to Shinder, diagnosis took nine minutes; the fix was to move four cables to a spare switch in the cabinet. Dispatch did not resume until later that afternoon. The delay was attributed to three practical obstacles: the contract tracked an engineer’s arrival rather than restored service, the cabinet was locked and its keyholder was away, and the spare switch had not been configured. Read Shinder’s account on DEV Community.
Why fast diagnosis did not mean fast restoration
Remote diagnosis and onsite restoration are different parts of incident response. An engineer may identify the likely fault quickly, but a repair that requires access to a cabinet or moving cables still depends on someone being physically present and authorized to act. If that person cannot get into the room, or the replacement equipment is not ready, the diagnosis does not restore the service.
Free tools Windows power users keep installed
One-click scans. No signup required.
The depot incident also shows why the wording of a support metric matters. An arrival-time target can be met even if the business remains unable to dispatch. A service-restoration target instead focuses attention on the outcome the site needs. The account does not provide contract terms or measured comparisons, so it illustrates the distinction rather than proving which target is best for every organization.
What changed after the incident
Shinder says the organization introduced several measures to give remote engineers a safe path from diagnosis to action:
- A named contact at every site: someone local could receive instructions and help with physical access.
- A laminated instruction card: it listed six actions a site contact could perform under guidance. Examples in the article include photographing a panel, reading a label, power cycling a specified device, moving a cable between colour-coded ports, checking a socket, and unlocking the room.
- Documented comms rooms: rooms were photographed and labelled so remote staff could identify equipment and connections.
- Controlled key access: a key safe was installed, with its code held by the service desk.
- Configured spare equipment: spare switches were prepared for the four largest depots rather than left unconfigured until an outage.
- Out-of-band access: a mobile console provided a way to reach management interfaces when the production circuit itself failed.
- An outcome-focused contract: the support agreement was rewritten to measure a working service.
These controls address different failure points. A local contact and access arrangement help make physical intervention possible; labels and photographs help remote staff guide it; a configured spare reduces work during the outage; and an independent management path can help when the ordinary circuit is unavailable. The article does not specify equipment models, compatibility requirements, or detailed security procedures for the key safe or mobile console.
What the account reports—and what it does not establish
Shinder reports that four of eleven site incidents since April were resolved by a supervisor using the instruction card within forty minutes. The year meant by “since April” is not stated. This is the author’s reported outcome, not independently audited evidence or a controlled comparison; it should not be treated as a general success rate for other organizations.
The account does not quantify salary savings, the cost of the changes, or the financial impact of delayed dispatch. It therefore cannot show whether two centralized engineering locations are less expensive overall than regional staffing. It does make the trade-off concrete: staffing savings do not themselves provide hands at a remote site, and readiness measures take planning and resources too.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to assess a centralized support model
Use the incident as a practical checklist for comparing support arrangements, rather than as a universal verdict on centralization:
- Check the service outcome: determine whether the support clock measures arrival, response, or restored service—and whether that matches the operational impact of an outage.
- Map physical dependencies: identify which common fixes require a person to unlock a room, inspect a panel, check power, or move a cable.
- Assign bounded local actions: nominate a contact and document only actions that can be performed safely under clear instructions.
- Prepare for access and replacement: keep room documentation current, define controlled key access, and configure any spare equipment before it is needed.
- Test the failure path: confirm how engineers will reach management interfaces if the main network connection is down.
- Compare total operational trade-offs: weigh staffing and coverage costs against site readiness and the time required to restore service. Shinder’s article offers no independent cost-benefit analysis.
Central teams can provide expertise across many locations, but physical work still has a location. Shinder captures that limit in his closing line: “Centralising support removes the salary at the site. It does not remove the physics.”
Quick Recap
Best Value
- Quality material used to make all Pro force products
- Tested in the field and used in the toughest environments
- 100 percent designed in the USA
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Recommended Free Tools
