Page 1 of 2 12 LastLast
Results 1 to 20 of 37

Thread: The fragile cobweb of BA's computer network is nothing to be concerned about.

  1. #1
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default The fragile cobweb of BA's computer network is nothing to be concerned about.

    Nothing to see here. Nothing needs to be done. Definitely not some sort of regulatory requirement for the cash cow major airlines to invest in modern REDUNDANT networks. Definitely not that.

    http://edition.cnn.com/2017/05/27/wo...ure/index.html

  2. #2
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    3,964

    Default

    Unless you have a deep distrust that this somehow leads to bypassing of safety procedures...

    I concur that the fragility of the system and lack of backup are significant for the customer service and stock price departments.

    As we have established, the former is not a priority... the latter; however, definitely deserves a thread in an obscure discussion forum somewhere!
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  3. #3
    Member
    Join Date
    Jan 2009
    Posts
    111

    Default

    Quote Originally Posted by Evan View Post
    Nothing to see here. Nothing needs to be done. Definitely not some sort of regulatory requirement for the cash cow major airlines to invest in modern REDUNDANT networks. Definitely not that.

    http://edition.cnn.com/2017/05/27/wo...ure/index.html
    No one regulates the efficacy of running a business. To do so, MIGHT eliminate instances such as this, but at the huge expense of efficiency. This is not a safety issue, it is a business issue. It is no different than accounting forgetting to pay the fuel bill and suddenly someone stops the taps and the planes are all grounded for 5 hours while they sort it out.

    It also has nothing to do with Cobwebs or Network. Like the Delta incident, the information so far indicates it is a hardware failure which points to the opposite of a cobweb (i.e. it is very centralized and not decentralized like a web). Frankly, it just points to very poor risk management and fault tolerance to hardware failure at some key part of the system.

  4. #4
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default

    Quote Originally Posted by Schwartz View Post
    No one regulates the efficacy of running a business. To do so, MIGHT eliminate instances such as this, but at the huge expense of efficiency. This is not a safety issue, it is a business issue. It is no different than accounting forgetting to pay the fuel bill and suddenly someone stops the taps and the planes are all grounded for 5 hours while they sort it out.
    Except that instead of five hours, it is two days of chaos affecting thousands of people who depend on BA as a public transportation service. Governments must do what they can to prevent such epic fails as part of the social contract.

    It also has nothing to do with Cobwebs or Network. Like the Delta incident, the information so far indicates it is a hardware failure which points to the opposite of a cobweb (i.e. it is very centralized and not decentralized like a web). Frankly, it just points to very poor risk management and fault tolerance to hardware failure at some key part of the system.
    Exactly. A single point of failure that (perhaps through a cascade of subsequent failures) brings the entire thing crashing down. Imagine an A380 designed with a single engine. That's what we have here.

  5. #5
    Senior Member Gabriel's Avatar
    Join Date
    Jan 2008
    Location
    Buenos Aires - Argentina
    Posts
    5,833

    Default

    Quote Originally Posted by Evan View Post
    A single point of failure that (perhaps through a cascade of subsequent failures) brings the entire thing crashing down. Imagine an A380 designed with a single engine. That's what we have here.
    Or rather, imagine a twin engine airplane capable of carrying 300+ pax across the Pacific ocean designed with a single point of catastrophic failure, in such a way that a catastrophic failure in one engine could induce a catastrophic failure in the only other engine.

    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
    --- Defend what you say with arguments, not by imposing your credentials ---

  6. #6
    Member ATLcrew's Avatar
    Join Date
    Sep 2013
    Posts
    576

    Default

    Quote Originally Posted by Evan View Post
    Except that instead of five hours, it is two days of chaos affecting thousands of people who depend on BA as a public transportation service. Governments must do what they can to prevent such epic fails as part of the social contract.
    Your trust in a benevolent (and presumably competent) government remains endearing as ever.

  7. #7
    Member
    Join Date
    Jan 2009
    Posts
    111

    Default

    Quote Originally Posted by Evan View Post
    Except that instead of five hours, it is two days of chaos affecting thousands of people who depend on BA as a public transportation service. Governments must do what they can to prevent such epic fails as part of the social contract.



    Exactly. A single point of failure that (perhaps through a cascade of subsequent failures) brings the entire thing crashing down. Imagine an A380 designed with a single engine. That's what we have here.
    Except it isn't the same as a 380 flying with one engine. No one dies when the plane doesn't take off from the ground. People only get inconvenienced, and that makes all the difference. Same way they get inconvenienced when a volcano erupts, or a hurricane raps the east coast. Five hours of no flights from lack of fuel would cause an equally damaging cascade effect on schedules for days.

    They have decided the cost of the extra redundancy wasn't worth the damage from an incident like this. Calling for regulation of this is ridiculous. You will end up with computer systems like the ones on the space shuttle.

  8. #8
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default

    Quote Originally Posted by Schwartz View Post
    Except it isn't the same as a 380 flying with one engine. No one dies when the plane doesn't take off from the ground. People only get inconvenienced, and that makes all the difference. Same way they get inconvenienced when a volcano erupts, or a hurricane raps the east coast. Five hours of no flights from lack of fuel would cause an equally damaging cascade effect on schedules for days.
    I object to the words 'only' and 'inconvenienced' there. If you had to sleep on the floor of an airport for two days because an airline can't provide fault tolerance in their networks, you might object to them as well. And let's not confuse a manmade clusterfuck for a natural disaster. This disaster was entriely preventable. As the next one will be. And the next one and then next one...

  9. #9
    Member
    Join Date
    Jan 2009
    Posts
    111

    Default

    Quote Originally Posted by Evan View Post
    I object to the words 'only' and 'inconvenienced' there. If you had to sleep on the floor of an airport for two days because an airline can't provide fault tolerance in their networks, you might object to them as well. And let's not confuse a manmade clusterfuck for a natural disaster. This disaster was entriely preventable. As the next one will be. And the next one and then next one...
    Seriously? I suppose I should object to a comparison of a plane crashing killing all aboard to a couple days sleeping in an airport instead? Please. People can always choose to go to a hotel or home. I remember getting caught at airports in NYC and making a quick decision that I was going to leave and find somewhere to sleep. They don't have to stay there for 2 days. These things are always a risk when flying.

    As for preventable? Many problems and even deaths (automobile comes to mind) are preventable in life. Life most other things, it is purely a matter of cost and ROI. Again, these are business decisions, nothing more.

  10. #10
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default

    Quote Originally Posted by Schwartz View Post
    Please. People can always choose to go to a hotel or home.
    I want to live in your world.

    Meanwhile, day three of clusterfuck. IT guys still trying to make sense of the cobweb. Not only do we still have zero fault tolerance, we also have fix intolerance. What a lovely mess waiting to happen this was.

    Airlines can always choose to build robust networks.

  11. #11
    Member
    Join Date
    Jan 2009
    Posts
    111

    Default

    Quote Originally Posted by Evan View Post
    I want to live in your world.

    Meanwhile, day three of clusterfuck. IT guys still trying to make sense of the cobweb. Not only do we still have zero fault tolerance, we also have fix intolerance. What a lovely mess waiting to happen this was.

    Airlines can always choose to build robust networks.
    Last time I checked we lived in the same world, where corporations trade off customer service with cost ALL THE TIME.

  12. #12
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    3,964

    Default

    Quote Originally Posted by Schwartz View Post
    Last time I checked we lived in the same world, where corporations trade off customer service with cost ALL THE TIME.
    I only side with Evan that its amazing when this happens... it has to be expensive and one would think 'worth' avoiding with some pretty robust efforts and back ups.

    A whole days profit and expensive rectifications...
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  13. #13
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default

    Quote Originally Posted by 3WE View Post
    I only side with Evan that its amazing when this happens... it has to be expensive and one would think 'worth' avoiding with some pretty robust efforts and back ups.

    A whole days profit and expensive rectifications...
    Well, I also think it's a matter of ethics and responsibility to your society (the one that makes it possible for you to prosper in the first place). It's just not cricket to leave people in a state of chaos that could have been prevented by a little less bottom-line and a little more high-mindedness. But all that is so quaint in an age where the word "business" is some sort of ubiquitously acceptable excuse for not having a shred of foresight or concern for the poor weary traveler who—no—does not often have the option of a hotel or home, but must rather wait around stranded in an absurd state of limbo for days in everyone's least favorite place to lay one's head, the modern airport.

    But you are certainly right 3WE, the expense of this debarcle will outstrip the cost of having prevented it. The problem is, this is not the result of practical and visionary cost/benefit thinking. This is push-it-down-the-road-cross-that-bridge-when-we-come-to-it-careening-headlong-into-quarterly-performance-stock-valuation-because-the-largesse-of-my-compensation-is-determined-by-share-price-valuation thinking.

    All of which could be avoided by sensible regulatory standards for the network fault tolerance of airline operations that societies entirely depend on to function.

    And lastly, I feel I'll need to repeat this until the tragic day I'm proven right: any element of chaos which imposes stress and pressure on commercial aviation IS, through an unpredictable chain of events, A SAFETY RISK.

  14. #14
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    3,964

    Default

    Quote Originally Posted by Evan View Post
    ...And lastly, I feel I'll need to repeat this until the tragic day I'm proven right: any element of chaos which imposes stress and pressure on commercial aviation IS, through an unpredictable chain of events, A SAFETY RISK.
    Of course, the pilot getting out of bed and going to work may have an even higher correlation to crashes than the link you keep pushing... just a thought.
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  15. #15
    Member
    Join Date
    Jan 2009
    Posts
    111

    Default

    Quote Originally Posted by Evan View Post
    All of which could be avoided by sensible regulatory standards for the network fault tolerance of airline operations that societies entirely depend on to function.
    Not a chance. You have absolutely no idea about how impractical what you are suggesting is. Also, you are very quick to jump to conclusions without having any real facts to work with. What would others say if you stated the cause of the crash before you even had a single fact about the events leading to the crash? Apparently, that's OK in this case? I am guessing, but I think there is a good chance making a system like this far more fault tolerant would require a complete overhaul of all of their systems. That is extremely expensive. I am sure that if we look at this single failure, there is a cheaper solution for it. However, I'll also bet there are 100 other similar vulnerabilities lying there which is something only a complete overhaul and re-architecture could address. I'll even go further to suggest, they should not make the system resilient to single points of failure. Instead, they should accept the failure and allow the system to recover quickly when it does fail.

    Well, I also think it's a matter of ethics and responsibility to your society (the one that makes it possible for you to prosper in the first place). It's just not cricket to leave people in a state of chaos that could have been prevented by a little less bottom-line and a little more high-mindedness. But all that is so quaint in an age where the word "business" is some sort of ubiquitously acceptable excuse for not having a shred of foresight or concern for the poor weary traveler who—no—does not often have the option of a hotel or home, but must rather wait around stranded in an absurd state of limbo for days in everyone's least favorite place to lay one's head, the modern airport.
    I think the real problem is the lack of competition which is driven by cost saving through consolidation, and enabled by lack of competition watchdogs by government. If your airline is leaving you stranded, go use a different one. They will invest more heavily in their systems really quick if their monopoly was really at risk.

    And lastly, I feel I'll need to repeat this until the tragic day I'm proven right: any element of chaos which imposes stress and pressure on commercial aviation IS, through an unpredictable chain of events, A SAFETY RISK.
    You can always claim to be proven right on this because it is flawed logic at the core. If Commercial Aviation is not resilient to schedule pressure, then that safety problem was caused by the system, not the source of the stress.

  16. #16
    Member ATLcrew's Avatar
    Join Date
    Sep 2013
    Posts
    576

    Default

    Quote Originally Posted by Evan View Post
    But you are certainly right 3WE, the expense of this debarcle will outstrip the cost of having prevented it. The problem is, this is not the result of practical and visionary cost/benefit thinking. This is push-it-down-the-road-cross-that-bridge-when-we-come-to-it-careening-headlong-into-quarterly-performance-stock-valuation-because-the-largesse-of-my-compensation-is-determined-by-share-price-valuation thinking. All of which could be avoided by sensible regulatory standards for the network fault tolerance of airline operations that societies entirely depend on to function.
    With the possible exception of healthcare (which is in its own endless clusterbleep), I don't know of an industry more regulated than commercial aviation. Nevertheless, these sorts of things continue to happen. Is it at all possible that more regulation is not the answer?

  17. #17
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default

    Quote Originally Posted by ATLcrew View Post
    With the possible exception of healthcare (which is in its own endless clusterbleep), I don't know of an industry more regulated than commercial aviation. Nevertheless, these sorts of things continue to happen. Is it at all possible that more regulation is not the answer?
    Well, the best answer is ethics, but that ship has sailed.

    The next best answer is market forces, competition, the threat of lost revenue from downtime driving preventative investment. But, thanks to corrupted regulatory bodies, we now have virtual monopolies. So you can forget about that as well.

    The only option I see left is to impose the rule of law. Impose a set standard of provable resiliency and contingency for failure to the computer networks that have become so essential to safe, functional air travel. Pay to play, basically. And yes, they would pay, because they have the money and they want to play.

  18. #18
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,323

    Default

    Quote Originally Posted by Schwartz View Post
    Not a chance. You have absolutely no idea about how impractical what you are suggesting is.
    You've got to be kidding me. What could be more impractical than running your airline operations on a wing-and-a-prayer network and incurring 3+ days of chaos at any given moment? You want to talk about practicality?

    I am guessing, but I think there is a good chance making a system like this far more fault tolerant would require a complete overhaul of all of their systems.
    Yes

    That is extremely expensive.
    Yes, but an inevitable expense that will only become more expensive the longer you push it down the road.

    I am sure that if we look at this single failure, there is a cheaper solution for it.
    Cheaper in the short term or cheaper overall? Certainly you can keep sticking cheap patches on the thing, like an aging bicycle innertube. But what about the cost of ongoing disruptions and damage to your brand? Isn't it better to spend more on a new, more robust, less-puncture-prone innertube? Isn't it cheaper to spend your time generating revenue rather than patching and apologizing?

    However, I'll also bet there are 100 other similar vulnerabilities lying there which is something only a complete overhaul and re-architecture could address. I'll even go further to suggest, they should not make the system resilient to single points of failure. Instead, they should accept the failure and allow the system to recover quickly when it does fail.
    Agreed! The issue isn't about making something fail-proof, it is about making it fail-passive (or fail-operational). That's all I'm getting at. If the booking system goes down for 30 mins, that's manageable. But to do this, you have to plan it out architecturally and have control and supervision of every mission-critical aspect. You have to have contingencies in place and a means to avoid failure cascades. Instead of a house of cards you need to have a house where any card is able to fall without causing the entire thing to come crashing down.

    But you can't have that when you cobble together networks from a jumble of third-party legacy components and outsourced IT. Until you get rid of that and build something modern and manageable, all you can do is patch and pray...

    If Commercial Aviation is not resilient to schedule pressure, then that safety problem was caused by the system, not the source of the stress.
    Unless you've had your head in the sand all these years, you are aware of the many accidents caused by schedule pressures, manifested in get-there-itis and fatigue-inducing duty rosters. The industry is indeed resilient to these pressures but with notable exceptions which are quite notable for their mass fatalities and impact craters. Yes, the source of the stress IS the system because the system creates and tolerates such pressures, either deliberately through profit seeking or unintentionally through neglect.

    There is only one word to categorize what has happen to BA these past days: neglect.

    If your airline is leaving you stranded, go use a different one.
    Again, I want to live in this world of yours. I'm sure all the stranded passengers from the past three days would as well.

  19. #19
    Member ATLcrew's Avatar
    Join Date
    Sep 2013
    Posts
    576

    Default

    Quote Originally Posted by Evan View Post
    Well, the best answer is ethics, but that ship has sailed.

    The next best answer is market forces, competition, the threat of lost revenue from downtime driving preventative investment. But, thanks to corrupted regulatory bodies, we now have virtual monopolies. So you can forget about that as well.

    The only option I see left is to impose the rule of law. Impose a set standard of provable resiliency and contingency for failure to the computer networks that have become so essential to safe, functional air travel. Pay to play, basically. And yes, they would pay, because they have the money and they want to play.
    That's not what I asked.

  20. #20
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    3,964

    Default

    Quote Originally Posted by ATLcrew View Post
    ...Your trust in a benevolent (and presumably competent) government remains endearing as ever....Is it at all possible that more regulation is not the answer?...
    In a fantasy world, Evan is correct.

    In the real world, I have to renew my License plates periodically.
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •