Krueger says that this is problematic because it can have spikes in user traffic and interest at random. Krueger compares Black Friday to online retail sites, where companies can prepare for large traffic events with some certainty. Krueger says that there's a chance of a Black Friday on any given day. A news event can have a big impact on the discussion. Krueger says that it is difficult to lay off up to 80% of your SREs, but MIT Technology Review has not been able to confirm it. The engineer thought the percentage was plausible.
The current engineer doesn't see a way out of the issue other than reversing the layoffs. Things will break if we push at a rapid pace. There is no way to avoid that. We are accumulating technical debt more quickly than before.
He shows a future where issues pile up as the number of maintenance tasks increases. Things will fall apart. There will be more broken things. Things are going to break for a long time. He says things will be broken more severely. Everything will compound until it becomes useless.
The engineer says that the collapse is some time off, but the telltale signs of process rot are already there. The engineer says that the first thing they need to do is fix bugs in whatever part of the client they are using. As the fixes are being delayed, things will accumulate until people stop giving up.
Krueger believes that we will start to see a greater number of accounts coming into and out of existence because of the lack of loading. Krueger expects anything that's writing data on the back end to have slowness, timeouts, and a lot more subtle types of failure conditions. They are often more harmful. They take a lot more effort to get there. It's going to be a problem if you don't have enough engineers.
There are signs that this is already happening. Krueger says that they're seeing cut-down versions of failsafes that the platform can fall back on if they're needed.
There will be significant outages on the horizon, thanks in part to Musk's cost-cutting drive to reduceTwitter's cloud computing server load as an attempt to claw back up to $3 million a day in infrastructure costs The project came from Musk's war room. One of the sources said the plans were "delusional" and that the risk of poorer capacity and availability seemed to be a logical conclusion.
There is no longer an institutional knowledge to fix issues when they arise. A lot of the people who were leaving after Friday have been there for a long time. Years of knowledge about how its systems worked disappeared when those individuals walked out of the office. Some within the company argue that the knowledge base is too concentrated in the minds of a small group of programmers.
All Rights Reserved © 2024
