One size fits all has benefits, but when you fail to talk to the customer and users involved that adopt your technology, then you're wasting everyone's time. We found ourselves spinning our wheels with a contractor "standardizing" on all the gear being deployed for a customer network upgrade. Had the contractor exercised the use of "site survey" to fully understand the customer configuration and business needs; adding a router in front of another router at multiple locations wouldn’t be an issue. Never mind the complexity, and the probability that more finger pointing will evolve with this configuration--the costs to maintain will increase and the customer still doesn’t get what they really wanted.
The goal to simplify network management is well intentioned but it failed to accommodate the provision for a viable alternative route (link) to handle all traffic--data, voice and video. When the primary link fails (not if it fails), then the secondary link must support the customer's traffic. 4G and 4G LTE that was provided simply won't support this particular customer's volume of traffic without significant disruption to the practice.
The contracted and carrier-charged provisioning manages the network infrastructure and excludes outside parties from touching the routers. No problem, so long as we have accountability. Now we, including the supporting IT contractor, will have access to Voice Quality Monitoring metrics and packet capture files, something that the contractor overlooked. A carrier isn't the same thing as a managed service provider, but a carrier will monetize the opportunity to provide managed services. I'm not splitting hairs here, because managed service providers aren't normally building their business models based upon rates and tariffs approved by their State PUC or PSC. Then, with extended outages, managed service providers' responsibility for costs often kicks in with penalties, while the prorated cost of the regulated service from the carrier isn't substantial. But for other enterprises maybe the lines are blurred between service provider and carrier, and the details remain within the contract.
When we requested a trouble history for the customer we met more resistance and the objection was, "Why bother? It's a new carrier." This is a huge presumption, since not all issues reported are carrier-caused. Our customer asked for help since they were getting the same routine from the help desk. Thus enters the packet trace to isolate why users are getting knocked off of secure sessions (https). We found two key causes. First, all the routers were configured with a default timer for https traffic limiting sessions to expire after 10 minutes; also, on the virtualized server in the data center, another timer for inactivity was set too low. Getting back to knowing and understanding the customer helps, and we knew that their workflow was laced with challenges that slow down and hinder the users' work.
In a prior post, UC, Meet Time and Motion, I mentioned the customer workflow designed by delivered technology isn't always the most efficient. Entering patient data to collect on claims is basically tedious and time consuming. Using dual monitors because some of the workflow snags experienced by the users won't go away equips the users with a better tool that eases some pain in dealing with those snags. Similarly in the current instance, resolving the timer issues in the router and the security settings for inactivity are granular details that had contributed to the overall inefficiency and negative user experience.
Their network and technology in place for the past five years with numerous calls by the users to the help desk never netted them a guy to show up with a laptop to do a packet trace and to dig into what was going on to cause so much user frustration. But the key point about this customer is their voice traffic is their bread and butter, and to ignore the importance of the voice traffic led the contractor down the path of using a less than desirable backup route for backing up heavy volumes of traffic and selecting a router in the hopes of making it easier to manage just one model of routers across the network.
Analysis is good until it's myopic and laced with an ideal that disrupts the mission of the company. Knowing the customer and assessing their "needs" often leads those deploying technology down different paths. It doesn't mean "no compromise" but it also doesn't mean "shove it down the customer's throat" because you know better than the customer or user. So the message is, it's an imperfect world and how a customer walks down the technology path to accomplish what they want is more important than what you want to deploy or sell.
