Monthly Archives: January 2008

Some slices on ey00 were rebooted around 3am PST due to some corruption on the maintaned node. All slices were immediately brought back and everything is in working order.

Customers affected by this issue were only those utilizing the Coyote load balancers.  The downtime was sporadic and lasted approximately 30 minutes for those affected.  This issues was resolved by rebooting the load balancers.  If you were an affected customer and are still experiencing sporadic connectivity, please call the support line.

The interruption to ey00 environments appears to be resolved, and sites are reachable again. We are continuing to monitor, and are investigating what caused the issue. Our initial assessment is that the issue affected customers on the Coyote load balancers.

(edited to reflect load balancer, not routing issue)

We are currently experiencing an issue connecting to ey00 slices. We are investigating, and will report more soon.

(edited to reflect that it was a load balancer issue, not a routing issue)

A node on ey00 required a reboot at 4:30am PST. The downtime lasted 5 minutes but no major outages were detected. The cause was memory starvation. Administrators are working to solve the problem.

We’ve repaired the node outage and all slices are coming back up. The problem was caused by a temporary disruption of the clustering state on the affected node.

A node on EY00 went down at 2:40 PM PST. The server is being power cycled. Affected customers will be informed momentarily and more status will be posted here once the cause is found.

Some intermittent routing problems affected all clusters around 4:30 PM PST. The problem was resolved within 5 minutes. If status is available from Herkales, it will be posted here.

There were intermittent connectivity issues to clusters colocated at Herakles Data starting at approximately 2:58am PST. Problems were resolved in about five minutes.

All indications point to some sort of upstream routing event.

The interruption was brief and intermittent. Connectivity is currently restored. We have contacted the appropriate parties and expect an official response is forthcoming.