Global IT outage: How one software update could go so wrong | DW News

It's the sheer human arrogance and naivety of large tech companies that failed to do simple verification on their work resulting in such low tech mistakes

Just to remind people, in january this ceo was bragging to investors about reducing workforce expenses, and using more contract workers from india. Since cloudstrike went public, all it has been doing is trying to squeeze out profits for investors

It bears repeating: This is a CrowdStrike issue, not a Microsoft or Windows issue. The blame lies with CrowdStrike.

This was what was feared back in the days of Y2K

deeply sorry? ask how much it costs now.

At least flight were cancelled as their emissions...

How do we, as europeans, once again leave critical infrastructure in the hands of foreign companies...

The blame should be shared by providers as well as the using companies that all allowed an untested update to be rolled out to their systems. Society essential companies and providers such as hospitals and airports should have acceptance testing as a mandatory step of their update procedures. Note that not all Windows users faced this issue, but only those that accepted this update without prior testing. Only complacent companies that trusted automatic updates to never go wrong got to learn this expensive lesson.

many 'experts' say that software testing in a special environment is industry standard. That was true 10 yrs ago, but many CEOs have dumped that as old fashioned, and expensive. Instead they just threaten their employees to 'be careful' or you're fired. That's the only way an errant file in an update could get past 'standard' testing practices. Cloud servers run hundreds of applications from various companies and have strict memory access security policies to protect one app affecting or 'peeking' into the memory of another on the same server. It does it by assigning strict memory allocation rules, and should an app go outside of those rules, the app is shut down for investigation. What happened the security application update itself contained a file that was flagged as accessing memory out of bounds and the security app was itself shut down after the update with the errant file, blocking any and all transactions that attempt to connect to it. I doubt there was industry standard testing applied, otherwise the errant file should have been detected just as quickly as the updated version was.

we have entered the age of incompetence

2:20 Im sure that he’s deeply „sorry”

The did strike the crowd.. my whole travel plan is jeopardised

Windows is like the hot mail of operating systems 😂

The problem is not only because of update, it is more, because of updates everywhere, on all computers at the same time. When you can not more, like earlier turn off updates on your comp, or install only well tested updates, a little bit later.

It is not end, outsourced non-competencies pose more threats in the future. The software update did not trigger "clustered" environment back to previous installation of system to run uninterruptedly if update fails. More to come to learn.

Watched many interviews regarding this outage and this person they interviewed was the most knowledgeable.

CS is not traditional antivirus software. It's a network monitoring solution sending data from sensors/agents directly to the CS-managed customer endpoint. The solution focuses on endpoint security, detecting early signs of malware/cyber attacks by analyzing sensor data. Friday's CS update aimed to improve detection but caused chaos instead. 😂😂😂

a lot of cutting corners this year for big companies

Many years ago, it'd already been said by some computer experts that the window OS system at that time hadn't been stable enough, especially when it was compared with other then similar competing OS system like say UNIX/ORACLE. The UNIX/ORACLE system was more commonly used by the commercial banks in those past years, simply because those banks' computer system network could never have been down at any time for any excuse in any way, for very obvious safety reasons. Given this global IT outage, maybe the above view still holds today. (I think Apple's OS computer system has had UNIX as one of its basic inner core.)

THE WORLD SHOULD NOT DEPENDS ON ONE SYSTEM ONLY!