Wrong configuration

In the 1940s, Thomas J. Watson predicted that the world needed no more than five computers. In 1969, there were over 69,000 mainframe computers in the US. In Germany, there were over 8,000 in 1971. Over forty years ago, personal computers replaced the green host screens. Nowadays applications run in the cloud. We assume that the network runs stable and that all applications are accessible at all times. On October 4th 2021, 9 am PDT, Facebook, Instagram and WhatsApp went down for six hours. After hours, the world learned that a wrong configuration was the reason for the outage.

No doubt it was a matter of some wrong settings. However, we learn nothing about the reasons behind. The decisive factor is assuming the responsibility. However, there is always room for interpretation – if not for conspiracy theories. Wrong settings have many possible causes.

  • Operating errors
    There are no man-made things that are error-free. It may be due to the program code or the settings of a server or the adjustments for the Internet. It is often not the original error that appears, but only effects that have to be traced back. Meanwhile, the impacts spread. This results in many failures in a short period of time, which have to be rolled back eventually. In any case, it were wrong settings.
  • Sabotage
    An operating error happens without intention. When someone deliberately causes a breakdown, the malfunction feels initially like an operating error. However, the bad intentions lead to the fact that the saboteurs prepare their acts well in advance and obfuscate them accordingly. It can take some time until the actors are debunked. To what extent the real reasons will ever become known is an open question. In any case, it were wrong settings.
  • Malfunctions
    Failures quickly lead to hazardous incidents, which cause outage, which cause interruptions, and so on. Since errors spread at the speed of light, not only does the original defect need to be fixed, but so do all subsequent ones. Once the servers are no longer recognized, it takes hours to restore the original conditions. In any case, it were wrong settings.
  • Failed integration
    When complex IT systems converge, unforeseen difficulties quickly arise. In order to get rid of them, parameters are readjusted. This can then lead to aftereffects that build up. When systems fail, it takes time to modify the settings and restart the systems. In any case, it were wrong settings.
  • Security leaks
    Incorrect settings lead to open entry points that are used by hackers. It does not matter whether they pursue private, commercial or political intentions. Every program has a security hole that can be exploited with the right code. The fact that the backdoors are open is only detected by the operators after an incident occurred. However, it is important for us users to know if our data is safe. In any case, it were wrong settings.

Bottom line: Facebook has reacted professionally by taking responsibility. Yet, it remains open what is behind the incorrect configuration. You have to live with simple operating errors, because they are human – even after good training reduces the number. Saboteurs are already hard to find – despite the many log files. Malfunctions are fate, as they result from the complex interaction of IT systems and modules and no one can foresee them. When systems are upgraded to a new version or get harmonized, unpredictable errors occur. Security weaknesses are particularly unpleasant when personal data is stored. The actual reasons are only understood by experts and they remain in the dark to conceal the own weaknesses. The quick announcement that it was a wrong configuration of the provider is the only way to react responsibly – even if we don’t really know, what it was.