Botched network maintenance has been blamed for a huge crash at Saleforce’s data centres, taking out customers’ CRM and data services across the US and Europe.
Seven of Salesforce’s 17 North American instances and two out of four in EMEA went down. Servers started dropping out just before 2am UTC on 15 November, with outages varying for individual instances from two to three hours.
A message on the Salesforce system status page updating customers fingered planned network maintenance by engineers as causing the crash.
The message told customers: “The salesforce.com team is investigating the root cause of this issue. The preliminary findings point to planned maintenance in the network tier.”
As of late Friday, the status page still showed red icons that warned of service disruption with an accompanying message that gave details.
The Reg contacted Salesforce on Friday daytime for the actual cause and to find out how it was fixed, but the company declined to comment.
Rather, two spokespeople for the cloudy company simply referred us back to the service status page.
That’ll come as little comfort for Salesforce customers who were left in the lurch and who took to Twitter to vent their anger.
Tweeps complained that their entire businesses had been left unable to function because they couldn’t get their CRM apps or data.
Others demanded to know why Salesforce wasn’t communicating more on Twitter.
It’s ironic given Salesforce puffs on so much about socially enabling its own customers’ businesses. Its chief executive touts his services as something allowing stodgy old firms to talk to and engage with their customers through social media, allowing firms to mine users for new ideas.
Embarrassingly, the crash comes ahead of the company’s annual Dreamforce conference, scheduled for next week in San Francisco, California.
There, Salesforce CEO Marc Benioff will try to convince the world they should bet their businesses on the public cloud and on Salesforce, despite the latter having being felled by something as mundane as "routine, planned" network maintenance.
Inspiring words and vision in support of the public cloud will be delivered by Benioff’s peers and fellow top brass from Facebook and Yahoo!.
This morning's outage is Salesforce’s second crash this year, coming three months after data corruption took down servers in Europe for nine hours.
The company suffered two major outages in 2012, too, of between six and seven hours. Those were blamed on a fault in the storage tier and on a power failure at its west-coast data centre, which took down service for customers in that region. ®