Barclays customers were left without access to their accounts on Saturday thanks to a computer failure. Customers could not use ATMs or online banking thanks to a computer failure apparently caused by a power outage. The bank insisted the problem was fixed within 20 minutes - by about 1.50pm on Saturday. But some forum posters …
Gawd, I remember meetings, back during the dot com bubble when NCR were going to maintain kiosks for certain people.
I don't think they have any more of a clue now than they did then.
Fortunately, I did all my banking in the morning, so didn't suffer a problem.
ATM, cash machines
in my cadre we call them wall banks and I humbly offer it as the preferred term by El Reg. As it perfectly describes their function and friendliness.
It could equally apply
to the drones that work inside too.
But they should.....
Cue lots of idiots clamouring about "Disaster Recovery" and "they should have resiliance" when they know nothing about the situation.
Some errors just happen, and the knock on effect if bigger than the fault. e.g. ever been in a phantom traffic jam.
Even with the best will in the world, you cannot plan for every eventuality. Some are impossible to predict. As long as they learn something from this and prevent it (or similar things) happening again, then it's not a complete waste.
Reminds me of the time during a disaster recovery test when the guy running the test told me the PTB wanted me to document every possible problem we could encounter, and the resolution.
Well, that would be job security for the next 10 years or so...
Oh, I would love that breif
Stampeding Elephants crush datacentre.
Resolution: Surround datacentre with crack guard mouse platoon.
Resolution: Retain services of Bruce Willis
As you say, could keep you in easy employment for a decade!
ATMs and the associated back-end represent a tier 1 service (ie failure can lead to 'financial and reputational harm), so ought to be in the region of 4 or 5 nines availability. Sometimes, yes, sh*t does happen, but Barclays is not what it ought to be when it comes to resilience and failover.
Believe the Reuters story and you have an outage (partial outage, I'll wager) of, what, 8 hours? So that gives an annual uptime of 99.92%. That's not shocking and, yes, I'm assuming no more problems in the next year. Take it at the bank's 20 minute value and their uptime goes to 99.994%. Mind you, both these figures are more than acceptable - phone lines and terminal kit at retainlers are both far more likely to let me down, I suggest.
This gets a write-up because it happend on a Saturday - Sod and his laws, eh? If it had happend at 3am, the public would have been none the wiser!
Not 20 minutes
"The bank insisted the problem was fixed within 20 minutes - by about 1.50pm on Saturday"
Funny that it was happening to many people at 11:40 or before.
Uptime of 99.92% is acceptable in a single piece of equipment, but that is not good enough for the whole Dawkinsdamn system
Haven't banks heard of redundancy? IGMC