Webcentral's owner, Melbourne IT, has issued a statement stating the failure "impacted a small percentage" of its customer base, but iTnews understands the numbers actually affected could be several thousand.
It has been rumoured on the Whirlpool broadband forums that an IBM storage area network (SAN) failed, prompting WebCentral to transition some 20TB of data across to a new storage system.
Melbourne IT, given multiple opportunities to respond to these allegations, refused to confirm the exact cause of the failure.
A Sydney-based small business owner and now ex-WebCentral customer, who spoke to iTnews on condition of anonymity, said they had experienced 65 hours of downtime in a 106-hour period starting Friday.
"Our business generates 50 per cent of orders from web sales," the owner said.
"It's a lot of money we may or may not have lost."
The business owner described the customer service updates provided by WebCentral as "hopeless" and "vague".
It is understood that several statements were issued privately to affected customers through the mission control section of their administrative panels.
But customers claim the messages were vague on details such as the cause of the issue and when it would be resolved.
The SMB that iTnews spoke to said the service initially went down on the morning of Friday 20th March and did not get service restored until Sunday.
"It went down again on Monday morning and we got the same story from WebCentral - that only a handful of websites were affected," the owner said.
"It came back again at midday and then at 6.30 or 7pm Monday it went down again and came back the next day."
Melbourne IT issued a brief statement yesterday claiming the issue had been resolved on Saturday 21st March and that customers "had been restored over the weekend".
It said that "secondary effects of server corruption have been identified and are being resolved."
Further, it said that "due to redundancy within the storage systems the problem was isolated and impacted a small percentage of the customer base."
But the response on redundancy has left customers of its shared hosting service fuming.
"It annoyed the c**p out of us that Webcentral didn't have a fallback strategy in place to redirect our domain name to a page saying they were experiencing technical difficulties," the business owner that spoke to iTnews said.
"Instead our site just wasn't available. It gives the wrong sign to our customers.
"We're paying premium rates of $90 per month for shared hosting and we'd expect them to have some redundancy in place."
The owner also responded to suggestions on the Whirlpool broadband forums that it was the individual site owners' responsibilities to pay for a back-up host and redirect their URL to an apology that the site is experiencing difficulties.
"What you've got to understand about SMBs is that we're not IT experts," the business owner said.
"I assume if I pay for a premium service from supposedly Australia's largest and best hosting provider that they have systems in place to enable us or them to access this kind of fallback mechanism without us having to worry about having to go to another provider. That's what we're paying for them for."
Regardless, the owner said they were in the final stages of shifting their web site to a rival provider and that they would also now run a back-up option if redirection is needed in the future.
The owner said early experiences with the new host had been positive. "We're now getting page loads in the sub one-second range, whereas page loads on Webcentral were around 10 seconds.
"We found the response times on their servers were hopeless. I don't know what they're doing, but now we've [moved], it's super fast and so far, so good."
The owner predicted more businesses would jump ship following the outage. Several have spoken of their intentions to move on the Whirlpool forums.
"Anyone who experienced that [downtime] wouldn't [stay] around I'd think," the owner said.