Everhart, Glenn
From:	Steve Spires [sspires1@ford.com]
Sent:	Friday, July 10, 1998 7:07 AM
To:	Info-VAX@Mvb.Saic.Com
Subject:	Re: NT Meltdowns/Disasters; Documented Examples needed
Further ti this;

NatWest Bank has denied that the switching of Achi Racov from chief
 information technology officer to board adviser on technology resulted
from the
 public failure of its branch systems. 

 In June 300 NatWest cash machines were taken offline as a result of
problems
 with Microsoft's Windows NT operating system. The job change was
planned
 months ago, a bank spokesman said. Racov was closely associated with
the
 bank's decision to back NT. 

 Jim Mann has become director of group operations and technology in the
 fourth IT management re-structuring which the bank has carried out in
the past
 seven years. 

 Other banks are also using Windows NT but NatWest is believed to have
made
 an exceptionally big commitment to the software. Around 50,000
workstations
 at 2,000 UK sites are expected to be dependent on it. NT scalability
problems
 began to surface in 1996. 

IT staff at NatWest have identified a bug in Windows NT and an
overambitious
 software roll-out as the main culprits behind the failure of its retail
banking
 platform. A bug in the 3.51 version of Windows NT used by NatWest
caused the
 handle leak that led to 400 branches and 300 cash machines being taken
 offline a fortnight ago. 

 The bug, which has been acknowledged by Microsoft, relates to the Local
 Security Authority component in NT, which is responsible for giving
users secure
 access to the branch network. Details of the bug can be found in the
Microsoft
 KnowledgeBase at support.microsoft.com. The main symptom is that NT
runs
 out of network connections, preventing access to the network. Microsoft
has not
 publicly released a fix for the bug, although staff at NatWest said the
software
 giant has written a fix for the bank. 

 The problems at NatWest began on the evening of 23 June when the bank
 attempted to roll out a software patch for a client/server application
on the retail
 banking platform. Initially the roll-out was scheduled for 100 sites,
but IT staff at
 the bank said that the number had been increased to 300. 

 This, combined with the bug in NT 3.51, generated a traffic jam over
the
 branch network the following day. "We ran out of connections to the
servers,"
 said one of the bank's IT staff. NatWest has about 35,000 NT
workstations and
 plans to roll out a further 15,000. 


Steve Spires

From <http://www.computerweekly.co.uk> (requires registering - free)

Arne Vajhøj wrote:
> 
> > the NatWest fiasco in the UK, cited in other threads was a
> > fine example;   methinks there ought to be a few more similar
> > examples out there, ie huge cost overrruns, and lesser
> > reilability/performance.  etc. etc.
> 
> I have read it too.
> 
> It would be nice with some details. What was the problem
> exactly ? How did the story end ? etc.etc.
> 
> Arne