Dokploy/dokploy · error · Error

${sanitizeRegistryError(execError, response?.password)}

Error message

${sanitizeRegistryError(execError, response?.password)}

What it means

Inside updateRegistry, after the DB update a docker login is re-executed with the stored/updated credentials (execAsyncRemote for a bound server, execAsync for cloud-type registries). If the login command fails, a plain Error is thrown whose message is the sanitized docker error (password masked with ***). This inner Error is then caught by the outer catch and re-wrapped as a BAD_REQUEST TRPCError with the same message — so the message you ultimately see IS this sanitized docker output.

Source

Thrown at packages/server/src/services/registry.ts:154

		if (
			IS_CLOUD &&
			!registryData?.serverId &&
			registryData?.serverId !== "none"
		) {
			throw new TRPCError({
				code: "NOT_FOUND",
				message: "Select a server to add the registry",
			});
		}

		try {
			if (registryData?.serverId && registryData?.serverId !== "none") {
				await execAsyncRemote(registryData.serverId, loginCommand);
			} else if (response?.registryType === "cloud") {
				await execAsync(loginCommand);
			}
		} catch (execError) {
			throw new Error(sanitizeRegistryError(execError, response?.password));
		}

		return response;
	} catch (error) {
		const message =
			error instanceof TRPCError
				? error.message
				: error instanceof Error
					? error.message
					: "Error updating this registry";
		throw new TRPCError({
			code: "BAD_REQUEST",
			message,
		});
	}
};

export const findRegistryById = async (registryId: string) => {

View on GitHub (pinned to 546686ea35)

Solutions

1. Reproduce manually: echo <password> | docker login <registryUrl> -u <username> --password-stdin on the target server
2. Fix/regenerate the credentials and retry the update
3. Verify the remote server is reachable if serverId is set

# verify the new credentials before updating
echo "$NEW_PASSWORD" | docker login <registryUrl> -u "$USER" --password-stdin

const isLoginFailure = (e: unknown): e is TRPCError =>
  e instanceof TRPCError && e.code === "BAD_REQUEST" && /password|unauthorized|login/i.test(e.message);

try { await updateRegistry(registryId, data); } catch (e) { if (isLoginFailure(e)) { /* message is sanitized docker stderr; fix creds */ } throw e; }

When it happens

Trigger: Updating a registry with new credentials that are wrong/expired, the registry being unreachable, or the bound remote server being offline at update time.

Common situations: Rotated registry tokens not updated in Dokploy, typo in the new password, remote server down when the automatic re-login runs after an update.

Related errors

• BAD_REQUEST
• NOT_FOUND
• BAD_REQUEST
• UNAUTHORIZED
• UNAUTHORIZED

AI-assisted analysis of Dokploy/dokploy@546686ea35 (2026-08-27). Data as JSON: /api/errors/4b083ec5a7004406. Report an issue: GitHub.