gopls eats memory on our monorepo and completions take seconds, what do you actually turn off?

I measured this properly because I was about to buy a laptop over it. Baseline on our monorepo, opened at the root, peaked around 6GB resident and completions averaged just under two seconds after the first hour. Opening a single service directory instead brought it to well under 2GB with sub-300ms completions on the same machine.

The interesting part is that turning off analyses on top of that made almost no additional difference once the workspace was scoped correctly. Scope first, then tune. Tuning a badly scoped workspace is rearranging deck chairs.

Three settings that measurably helped on our repo, in the order they helped. Turn off the extra static analysis passes, which run on top of the type checking and are the most expensive optional work gopls does. Turn off the code lens features you never click, particularly the run-test-above-every-function ones. And disable automatic imports scanning of the whole module if your editor is doing that on every keystroke.

After those three, peak memory on our repo dropped by roughly a third and completion latency became tolerable again. The analysis passes were by far the largest single contributor.

I want to push back on the turn-off-the-analyses advice, because it is treating a symptom and you lose real value. In our case the actual culprit was a stale build cache and a generated code directory that nothing excluded, so gopls was type checking several hundred thousand lines of generated protobuf output on every change. Excluding the generated directories from the workspace and clearing the build cache fixed it more thoroughly than any setting did, and we kept the static checks. Look at what is actually in your 600k lines before you disable the tooling you presumably turned on for a reason.

Stop opening the repo root. Open the subdirectory you are actually working in as the workspace folder and gopls only builds the graph for what is reachable from there. On a monorepo that is usually the single biggest change available to you, and it costs nothing.

I made this worse for a year by running three extensions that each did their own analysis on save, all of which duplicated work gopls was already doing. Linting on save through a separate extension, a formatter extension that shelled out on every keystroke, and a test runner that watched the whole tree. Removing two of them and letting gopls do its job single-handedly halved my CPU and I have never missed either.

Scope the workspace, exclude generated code, drop the optional analyses, and keep the language server on a recent version. In that order, and stop after whichever one fixes it.