Black Wednesday: DXC hosting services wonky for almost 8 hours after core switches go rogue
TITSUP* at Wynyard, Doxford bit barns leave finance clients with some free time – sources
The dust is settling at DXC Technology following an outage at its Wynyard colocation site in England yesterday that hosts services for a bunch of FTSE 50 financial services clients, insiders have told us.
Primary hosting is done at DXC Wynyard with disaster recovery provided by the DXC Doxford data centre.
Before lunch at around 11:25BST, DXC lost all access to the circa-200 servers hosted in those sites and normal play couldn't resume for more than seven and a half hours.
"We discovered that our connectivity to the site was good as far as the DXC managed firewall between our environment and theirs [clients] but none of the systems were accessible from the internal WAN or from customers on the internet," a source close to the outage told us.
Services began to clamber back at 12:29BST but remained intermittent. Resident techies were then told the secondary switch was shutdown by DXC and its hardware vendor was engaged.
The primary and secondary core switches had played up, with both trying to be the primary. Service interruptions continued, a source told The Register.
The active switch continued to wobble so services failed over to the other partner devices, which had to be restarted. Around 16:00BST the services started to be restored and full services resumed at around 19:00BST.
A technical report is pending.
"All in all not the best day," one of our sources at a financial services client said.
El Reg asked DXC about the outage but the company declined to comment. ®
* Total inability to support usual processes
Sponsored: Beyond the Data Frontier