Salad logoSalad logo
Salad logoSalad logo
Quickstart
Troubleshooting
Degraded StateDegraded State error codesFailed to update container environmentFailed to start container environmentFailed to start workloadSalad cannot access your GPUInsufficient VRAM and high VRAM activityHigh GPU usage, temperature, or the check can't runGPU performance lowGPU power limit setContainer networking issuesProblem detecting hardwareYour GPU is temporarily disabled
Discord
TroubleshootingDegraded State

Failed to start workload

Fix the "Failed to start workload" Degraded State notification (errors -1, -4, -5, -7, -10), where the container environment runs but a job can't be downloaded or started.

What it means. Your container environment is up, but a job couldn't be downloaded (-4) or started (-1, -5, -7, -10). The error number tells you which half failed.

Error -4: the job couldn't be downloaded

The job's files come from Salad's image registries. A download that never finishes or keeps restarting is almost always a network path problem on the PC rather than the app. Go to Downloads slow or failing.

Errors -1, -5, -7, -10: the job downloaded but wouldn't start

By far the most common cause is the GPU driver. Container jobs reach your card through your normal Windows NVIDIA driver; if that driver is damaged, only half-updated, or temporarily lost by WSL, the job fails at start with "-1" or "-7" even though games run fine. The second most common cause is a program that already has the GPU.

Check whether the container environment can see your GPU

In PowerShell (no admin):

wsl -d salad-enterprise-linux --exec /usr/lib/wsl/lib/nvidia-smi

A table listing your card means the driver link works - skip to step 3. An error means the environment has lost the card: continue with step 2.

Restart the container environment, then reinstall the driver

Exit Salad from the tray icon, run wsl --shutdown in PowerShell, and start Salad again. If the error returns, do a clean reinstall of the NVIDIA driver: How to reinstall your GPU drivers (choose Perform a clean installation), then reboot.

Free the GPU

Close games, local AI tools (Ollama, LM Studio, Stable Diffusion), video editors and mining software outside Salad. If you have two dedicated GPUs, only one can earn; disable the other in Salad under Settings > Configure Hardware (why).

Docker Desktop: turn off Resource Saver

If Docker Desktop is installed, its Resource Saver setting pauses the WSL machinery Salad relies on. Open Docker Desktop > Settings > Resources and disable Resource Saver.

Update WSL and retry

As Administrator: wsl --update, then wsl --shutdown, then start Salad and leave it for 30-40 minutes so it can download and start a job.

Last resort: reinstall Salad

Uninstall Salad, reboot, reinstall from salad.com/download. Do this only after the steps above - it doesn't touch the driver, which is usually the cause.

Still stuck?

Send Salad Support the notification screenshot with the error number, the output of the nvidia-smi command from step 1, and your logs folder zipped (how to find it).

Edit on GitHub

Last updated on

Failed to start container environment

Fix Degraded State errors -2 and -3, where Salad's Linux container environment can't be imported into WSL or its service won't start.

Salad cannot access your GPU

Fix Degraded State errors 11, 12, 19, 21 and 22 ("Salad cannot access your GPU", "GPU drivers out of date"): the GPU isn't usable from inside the container environment.

On this page

Error -4: the job couldn't be downloadedErrors -1, -5, -7, -10: the job downloaded but wouldn't startCheck whether the container environment can see your GPURestart the container environment, then reinstall the driverFree the GPUDocker Desktop: turn off Resource SaverUpdate WSL and retryLast resort: reinstall SaladStill stuck?