4 ms·
Aside from the political stuff, this illustrates how a company can be reliant on another vendor. AWS was a single point of failure for Parler (along with a few
by parsimo2010 6y ago
Aside from the political stuff, this illustrates how a company can be reliant on another vendor. AWS was a single point of failure for Parler (along with a few other services like domain registration).
For a one man show in a garage (or a very early stage startup), it's probably okay to focus on features and not backend resiliency. But by the time you have a board of directors, someone should have asked how they would handle losing a core service. Especially in this case someone should have seen that their business was high risk- either explicitly because of their plan/model, or incidentally because of the crowd they attracted.
Maybe Parler had a plan for some things, but they clearly weren't ready for a setback at this level. I wonder how prepared other companies are for losses/denial of service like this. Could Ford's supply chain weather a vendor dropping out? Could Kellogg's cereal?
- TZubiri 6y agoIronically, a one man show in a garage would be more resistant to legal challenges than hosting on aws. If you want to take down my server you have to either issue a warrant to search and seize my home, or a warrant for each of my ISPs, or intervene at the IP block or DNS level. The outcries over freedom of speech were just pathetic from a technological standpoint, it's a bunch of web 4.0 developers and users who felt entitled to services that even a 90s computer enthusiast would be able to provide. On some level it's not just Parler's fault, the internet has become more complex, and maybe some ISPs and modems provide some roadblocks like shared IPs and obnoxious firewalls, but after months they are still unable to lift a physical building place computers on property they control. It's like they are either incompetent or purposely looking for refusal of service so they can martyrize themselves.