Backend Development 11 min read

Twitter Engineers Clash with Elon Musk Over Performance, Code Freeze, and Microservice Architecture

A heated dispute erupts between senior Twitter engineers and Elon Musk as the latter imposes a sweeping code freeze and criticizes the platform's microservice‑driven architecture, prompting debates over RPC counts, technical debt, and the impact on app performance across regions.

DevOps
DevOps
DevOps
Twitter Engineers Clash with Elon Musk Over Performance, Code Freeze, and Microservice Architecture

Source: 技术琐话 (Tech Tidbits) – Author: Tina

This dramatic showdown has drawn legendary software modeling experts, container specialists, and infrastructure leads to collectively oppose Elon Musk, expressing support for the Twitter engineers who dared to challenge him.

What caused Twitter to refresh slowly?

On a Monday morning (U.S. time), Twitter engineers were summoned to an emergency meeting where Musk ordered an immediate freeze on all production changes.

The freeze was unprecedented: engineers were told they could not write any code until further notice, except for emergency production fixes approved by a vice‑president and explicitly justified to Musk. In Slack, engineers expressed confusion, noting the lack of background information.

Meanwhile, Musk publicly apologized for Twitter’s “extremely slow” speed in many countries, claiming the app executes over a thousand inefficient batch RPCs when rendering the home timeline.

Twitter engineer Eric Frohnhoefer responded, defending his six‑year involvement with the Android client and disputing Musk’s claim.

Musk replied, “Then correct me, what is the correct number?”

Frohnhoefer argued that while his team had improved performance, the Android app still had significant room for optimization and that the number of RPCs was not the main performance culprit.

He identified three reasons for slowness: many rarely used features bloating the app, accumulated technical debt forcing trade‑offs between speed and functionality, and network latency.

He suggested large‑scale rewrites to eliminate a decade of technical debt and remove unused features.

When asked about the RPC count, Frohnhoefer bluntly said, “Zero. The app does not make RPC calls at all.”

Other engineers, such as Ben Leib and Sasha Solomon, publicly criticized Musk’s understanding of Twitter’s infrastructure, with Solomon mocking Musk’s knowledge of GraphQL.

Veteran software architect Grady Booch, co‑creator of UML, warned that Musk lacks the ability to lead a globally‑scaled, software‑intensive organization.

Musk: I Can’t Argue With You, But I Can Fire You

According to Frohnhoefer, the Twitter app makes about 20 background requests on launch. Musk later claimed that up to 1,200 “microservices” are invoked when using the app, a figure Frohnhoefer disputed, stating the actual number is closer to 200.

Musk believes the excessive microservices cause slow refreshes and has been experimenting with shutting down services to see which are essential, even disabling the 2FA service.

While many large enterprises have bloated architectures, Twitter’s stack is well documented through numerous blogs and open‑source projects. Musk appears to be attempting a “pull‑the‑plug” test to see if performance improves by stripping down to core functionality.

The confrontation ended with Musk firing Frohnhoefer, who posted a screenshot of being removed from the corporate system.

Solomon, after being mocked for not knowing GraphQL, also announced her dismissal.

Infrastructure engineers shared a diagram showing that years ago a ride‑hailing service could require up to 4,000 services.

In Summary

Ten years ago, when Twitter tackled scalability and reliability, few open‑source tools existed. Engineers built world‑class storage, schedulers, and RPC frameworks, many of which were later open‑sourced.

Now Musk’s interventions have sparked widespread dissent among engineers, especially as performance varies dramatically across regions due to network conditions and device capabilities.

The reasons behind the code freeze remain unclear, though speculation points to Musk’s paranoia about potential sabotage from disgruntled staff.

Surveys of hundreds of Twitter employees on Blind indicate 89% doubt the company’s success under Musk’s leadership, while advertisers and high‑profile users are exiting the platform.

Engineers continue to mock Musk, coining terms like “crazing” in reference to former executives, turning the situation into a source of industry jokes.

Reference links:

https://www.theregister.com/2022/11/14/musk_twitter_rpc_spat/

https://twitter.com/elonmusk/status/1592177471654604800

https://twitter.com/ahmetb/status/1592189613774602242

https://www.platformer.news/p/elon-only-trusts-elon

https://www.theverge.com/2022/11/14/23458247/elon-musk-fires-engineer-correcting-twitter

https://news.ycombinator.com/item?id=33595520

https://news.ycombinator.com/item?id=33597677

# WAVE SUMMIT+2022 Deep Learning Developer Summit is coming! IDCF online zone still has seats – register now for exclusive gifts.

performanceMicroservicesBackend DevelopmentTwitterElon MuskCode Freeze
DevOps
Written by

DevOps

Share premium content and events on trends, applications, and practices in development efficiency, AI and related technologies. The IDCF International DevOps Coach Federation trains end‑to‑end development‑efficiency talent, linking high‑performance organizations and individuals to achieve excellence.

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.