Multi-Site Connectivity: Why Links Between Offices Fail

Multi-site connectivity means the links that join a firm’s offices, stores, plants and warehouses into one network. The sites then share applications and data. It works when each site stays online, reaches its cloud applications and has a link sized for today’s work. Branch network problems and office connectivity problems start when one of those three slips. A WAN is that wide-area network of links. This article covers why the links fail and when it hurts, and leaves building or choosing them aside. For the target state, see reliable connectivity for offices and sites.

Why Multi-Site Connectivity Fails: Five Causes

Branch network problems trace back to five plain reasons, and each one is easy to miss until a branch calls.

1. A site has one path. If a branch reaches the rest of the firm over a single line, one cut or one provider fault can stop the whole site. It does not matter what else works inside the building. A contingency planning guide for information systems from NIST lists redundant communications links among WAN measures. It advises that they follow separate physical paths, because one cable cut could otherwise disrupt both.
2. Traffic takes a detour through the center. In Signisys’s working view, some firms send a branch’s traffic to cloud applications through head office first. The user feels that delay as a slow application. In a hub design, traffic between two branches also passes through the central site. That can add delay between those branches.
3. Sizing dates from years ago. A circuit chosen for the way a site worked then can fall behind today’s video calls, cloud applications and backups.
4. Nobody watches the links. Without monitoring, users can report faults before the IT team sees them, and each report arrives as a vague complaint. The same guide describes monitoring software that can warn an administrator before users notice a problem. Putting that watch in place at each site is laid out in eight steps from a device list to alerts that reach a person.
5. Teams add new sites one at a time. Each branch gets a link ordered on its own, with no common plan for paths, sizing or monitoring. Every site ends up a little different.

What a Failing Link Costs a Business

Office connectivity problems cost a business in four places, and none of them needs a figure to make sense.

The site itself pays first. A store stops ringing up orders while the line is down. A warehouse stops printing dispatch notes, and a branch stops serving the customer at the counter.

Staff time is the second place. When a cloud application crawls at one office, people can wait, repeat work and find workarounds that nobody approved.

Engineering time follows. A fault can pull engineers off planned work. They then spend hours showing the provider that the problem sits on the line and not in the building.

Circuits make up the fourth. A firm can pay for a large link at a quiet site and a small one at a busy site, and not see it. Nobody can match the spend to what each site uses.

This article gives no cost figure. The right numbers sit in your own outage log, tickets and circuit list. The next two sections show where to read them.

When Multi-Site Connectivity Problems Show Up

Office connectivity problems reach the business at four moments, and each leaves a different trace in the records. Branch network problems tend to look like isolated complaints until someone lines the four moments up.

Each of these moments is about the link and how it performs. Records that point to the network itself are covered in how slowdowns on the network path go unnoticed. For the people side of branches, such as new staff and ordinary faults, see looking after IT across sites.

Signs Your Multi-Site Connectivity Needs Attention

Four signs tell you that office connectivity problems or branch network problems need a closer look. Each one reads from a record your firm already holds.

Start with the circuit list and count the sites that show one line from one provider. Then check the outage log for entries with no cause or no end time written down. Help desk tickets can show link complaints first filed against an application and later moved to the network team. Finally, the site register shows which links have no named owner or no monitoring entry. To score each link on those records, work through twelve checks that mark a link out of 24.

A sign supports a decision for the firm to make. A firm that sees office connectivity problems in these records must still weigh its sites, its budget and its risk. Where multi-site connectivity needs attention and the team has little time, an outside view can shorten the tracing.

What Stays Outside the Links

This article does not cover the security of the links, which branch office security risks takes up, or how to score a network. It also leaves out how to build the links or choose between link types.

SD-WAN is one way firms change how sites route traffic, and the SD-WAN page explains the term.

Which fix suits a firm’s branch network problems depends on its sites, trading hours and cloud applications. This article does not know those. It shows you the problem in records you already hold, before anyone spends money on a change. If the fix turns out to be a new kind of link, start with weighing a carrier circuit against software-steered links, site by site.

Signisys provides infrastructure and networking services, including enterprise networking for several sites, so weigh this view as a vendor’s.

If you would like help reading your own outage log, tickets and circuit list, a consultation starts there. Before any work starts, you and our team settle the scope.

Multi-Site Connectivity FAQ: Questions Answered