Skip to content

fix(server): one greedy agent process no longer takes down the whole server #5788

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
t3dotgg merged 1 commit into from
Aug 9, 2026
+17 −0
Merged

fix(server): one greedy agent process no longer takes down the whole server #5788

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 12 additions & 0 deletions apps/server/src/cloud/bootService.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,18 @@ it("keeps systemd pinned to the stable launcher rather than a versioned server",
expect(unit).not.toContain("versions/1.2.3");
});

it("survives the kernel OOM-killing a greedy agent child", () => {
const unit = BootService.renderBootServiceUnit({
nodePath: "/usr/bin/node",
launcherPath: "/home/theo/.t3/runtime/service-launcher.mjs",
baseDir: "/home/theo/.t3",
logPath: "/home/theo/.t3/userdata/logs/boot-service.log",
unitPath: "/home/theo/.config/systemd/user/t3code.service",
});

expect(unit).toContain("OOMPolicy=continue");
});

const makeHarness = Effect.fn("test.make_boot_service_harness")(function* (
platform: NodeJS.Platform = "linux",
usePinnedLauncher = false,
Expand Down
5 changes: 5 additions & 0 deletions apps/server/src/cloud/bootService.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -67,6 +67,11 @@ export function renderBootServiceUnit(plan: BootServicePlan): string {
// Let the launcher mark an explicit stop before it signals the server.
// systemd still SIGKILLs the whole cgroup if graceful shutdown times out.
"KillMode=mixed",
// Agent tool calls run as children of the server, so they share this cgroup.
// With the systemd default of OOMPolicy=stop, the kernel killing one greedy
// child stops the whole unit: the server, every live agent, and the user's
// connection. Keep running and let Restart=always cover the main process.
"OOMPolicy=continue",
"Restart=always",
"RestartSec=5",
`StandardOutput=append:${escapeSystemdSpecifiers(plan.logPath)}`,
Expand Down
Loading