Coles Group improves visibility across its SD-WAN

Coles Group has expanded its use of ThousandEyes software this year to instrument and keep tabs on the performance of an SD-WAN connecting its 1400-plus stores and other sites.

Coles Group's John Elhariry.

Network services team leader John Elhariry told Cisco Live in Melbourne that the retailer completed the SD-WAN rollout last year, and wanted the ability to be able to troubleshoot issues that may arise.

It has been using ThousandEyes - a network monitoring and digital experience assurance platform bought by Cisco in 2020 - for “a couple of years”, initially to monitor its “main core infrastructure” as well as the health of Internet-based services that it used.

Coles’ own network, from a hardware perspective, includes 20,000-plus access points, 5000 switches and 2000 routers.

In 2023, Coles expanded ThousandEyes to help it troubleshoot wi-fi issues in stores, and in 2024, it is using the software to keep tabs on its SD-WAN.

“This year we started to deploy ThousandEyes enterprise agents on the SD-WAN routers,” Elhariry said.

It set up and “started to run some tests, which turned out to be very beneficial.”

Elhariry said that ThousandEyes was used “a couple of months ago” to diagnose “a problem in the SD-WAN fabric”.

He indicated that prior to using ThousandEyes, the retailer had been uncertain about the scale of the problem. 

“At first glance [in ThousandEyes], we found an anomaly starting at a certain point in time, and that was impacting the whole state,” Elhariry said.

“That helped to say it’s more than a single store, service provider or a headend problem - it’s a broader problem.”

Elhariry cited another use of ThousandEyes to determine whether packet loss was occurring within its own network or once traffic passed to a third-party provider.

A dashboard, set up “a couple of months ago”, had shed light on the situation, and saved time that would otherwise have been spent “in conference calls troubleshooting what the problem was.”

In-house orchestration tool aids automation

Elhariry said that Coles’ network team broadly faced three challenges in its line of work.

The first challenge was lifecycle management - keeping its Cisco IOS software updated and hardware refreshed to stay in “compliance to the Cisco requirements”.

Other challenges are to “keep the network running” and ensure it can troubleshoot issues, and to have available capacity to support new opportunities and innovation.

“We found that automation and utilising modern observability tools will fit the first two challenges of lifecycle management and troubleshooting,” Elhariry said.

“If we save engineers’ time with the stuff that can be automated, or save them time in trying to troubleshoot stuff, then [they] have the time to do the innovation bit.”

On the automation side, Coles is making use of an in-house built orchestration tool that it calls My Portal, which integrates with several systems including Cisco Catalyst Center.

“For IOS lifecycle management, we utilise the functionality embedded in Catalyst Center to do the [IOS] upgrades, but we use My Portal to orchestrate this process end-to-end, so we don’t need to have an engineer up in the night doing the upgrade, the engineer just needs to schedule which store we need to upgrade, and then put in the time, and the tool will take care of it,” Elhariry said.

The tool also ensures that “pre- and post- [upgrade] checks are fulfilled.” 

“If everything is good, it sends an email to say [the upgrade is] completed,” Elhariry said.

“If there’s an issue, it automatically creates a ticket and then the after hours on-call teams will action it.”