Berichtje op 1 februari 2010 op de info pagina van One.Com. Nu weten we eindelijk wat er al een maand gaande is.
Network – 1 februari 2010 18:16 CET
Unfortunately, we have experienced instability on our network during the recent weeks, culminating in the days around Sunday January 31th, when many customers for brief periods could not access their website and email.
It is important to emphasize that no data or emails have been lost or will be lost as a result of these problems.
One.com has until now only used HP as the supplier of servers, but as of the summer 2009 we entered into an agreement with IBM regarding supply of blade servers. We have recently put a part of the new servers into service.
The server type has unfortunately caused problems for our service. to a large part of our customers this has meant periodical downtime for websites as well as many customers for short periods have not been able to read mail without experiencing clearly notable delays.
The hardware that has caused the problems has been delivered to One.com by one of IBM’s subcontractors. Today technicians from IBM are working closely together with our own technicians to solve the issues.
The technically detailed explanation is that an error occurred in the IBM-supplied switch module in which the blade servers are mounted. It is a brand new switch type from Blade Network Technologies. Periodically a heavy network noise is registered from the switch enclosure modules. The noise is disturbing the service, and occasionally spreading the problems to large parts of our network, making the whole network unavailable.
The solution we are about to implement is a total replacement of all the hardware, where we have seen even the slightest instability.
We apologize for any inconvenience and we hope for our customers’ understanding as we do everything possible to restore stability on our servers.