Salesforce.com customers experienced a loss of service for the third time in six weeks as the leading hosted applications provider was hit by another unplanned outage on January 30.
Salesforce.com has been hit by another outage.
The company website confirmed there had been system performance issues on the afternoon of Monday, January 30, 2006 causing intermittent access. Some customers said Salesforce.com servers were down from about 10am to 2pm Eastern Time.
According to reports, even when the company brought the service back up, its application programming interface (API) remained disabled for several more hours. This has wider implications because it means customers who interface their own applications with Salesforce.com using the Salesforce.com API would have been unable to connect to them.
The API issue compounds the problem at a time when the company is pushing forward with its ambition to transform itself into a platform provider and is touting the concept of the business web, which is its vision of a network of on-demand applications, services, components, and development work that will enable organizations to run their entire business under the on-demand model.
A database error on December 20, 2005 took the service down for about six hours, and on January 6, 2006 another outage affected the company’s Europe, Middle East, and Africa regions.
Customers say there have been minor problems since the release of the Winter 06 version, with some minor intermittent outages, degraded performance, and loss of access to some features.
Winter 06 is to a large extent a version 1.0-type application. At the combined Winter 06 and AppExchange launch, CEO Marc Benioff said the company and the application had undergone major changes to incorporate the idea of the business web and to enable the AppExchange.
Mr Benioff recently admitted that customers had experienced performance degradation in the preceding 90 days because the company was setting up the new data center, but said the December outage was not connected.
The December outage did not cause undue concern at the time because it was seen as a one-off, but three outages in such a short time period starts to look like a trend and will have used up goodwill. It will also raise questions about whether the company, which has 350,000 subscribers, is growing too fast and whether it has the necessary infrastructure.