When to Use This Guide
When you submit a job on Rescale, it passes through several states — Pending, Queued, Validating, Starting — before running. Brief waits are normal as the platform checks preconditions and provisions hardware. For details, see The Job Lifecycle and Job Statuses.
If your job has been in Pending for more than an hour or in Validating for more than 20 minutes with no sign of progress on the Status tab, it may be blocked. This guide walks you through a structured diagnostic flow.
Step 1 — Check the Status Tab and Status Reason
Open your job and navigate to the Status tab. The logs beneath the status badge is the most important diagnostic signal — it tells you why the job is in its current state.
Match your status reason to the corresponding section below:
Status Reason | Likely Cause | Section |
"Resource limitation" or "hardware resources are currently unavailable" | Cloud provider capacity exhaustion | Cloud Capacity Wait |
"vCPU Quota" | Provider-level quota limit | Provider Quota Limit |
"concurrent core limit of X" | Workspace or project core limit | Concurrent Core Limit |
No explicit reason, job remains Pending | License queue or platform issue | License Queue Block or Platform Issue |
"An error with the Rescale platform occurred" or "Error executing job" | Platform-side defect | Platform Issue |
If your status reason does not match any row, read through each section in order.
Cloud Capacity Wait
Status Reason: "Resource limitation"
The cloud provider (AWS, Azure, or GCP) temporarily lacks available instances of the type you requested in the selected region. This is an external constraint — your job configuration is correct.
Rescale automatically retries provisioning with increasing wait times. No billing accrues while the job waits.
What you can do (in order of effort):
Wait — Capacity often frees up within 30–60 minutes. Retries continue automatically.
Switch to an alternative coretype with similar specifications in Hardware Settings.
Expand region availability — Allow additional regions in Hardware Settings (need Admin approval and changes to be made by Rescale team)
Configure a Coretype Set — Enable automatic failover across multiple compatible hardware types. (need Admin approval)
Upgrade to On-Demand Priority (ODP) — Avoids spot-instance volatility with higher provisioning priority.
For a detailed walkthrough, see Understanding Cloud Provider Capacity Issues on Rescale.
Provider Quota Limit
Status Reason: mentions "vCPU Quota"
Your organization has reached the vCPU quota set by the underlying cloud provider (AWS, Azure, or GCP) on the account Rescale provisions into. This is separate from Rescale-level limits.
What you can do:
Contact Rescale Support to request a quota increase. Include your Job ID and the exact status reason text.
Quota increases require coordination with the cloud provider and may take time to process.
While waiting, try submitting a smaller job (fewer cores) or switching to a coretype under a separate quota category.
Concurrent Core Limit
Status Reason: mentions "concurrent core limit" — or no explicit reason but job remains Pending
Rescale workspaces and projects can enforce a Concurrent Core Limit (CCL) — a maximum number of cores running simultaneously. If starting your job would exceed this limit, the job queues until other jobs free up cores.
How to diagnose:
Check the Pending Jobs Queue view at the workspace or project level to see how many cores are currently active.
Ask your workspace or project administrator to review the CCL setting under Workspace Admin → Policies → Running Core Limits or the project's Max Concurrent Active Cores setting.
What you can do:
Wait for running jobs to complete. Your job starts automatically once there is room.
Ask your administrator to increase the limit if the workload requires it.
Reduce the core count on your job to fit within the available headroom.
Important: If a project's Concurrent Core Limit is set to 0, jobs can be submitted but will never start. This effectively pauses all execution within the project. If jobs consistently remain in Pending, verify the limit is set to a nonzero value.
For more on queue behavior and job reordering, see Understanding and Managing Job Queue Priority.
License Queue Block
No specific status reason, but the job's software requires a license
If the configured software requires a license and all available tokens are checked out by other jobs or on-premises users, the platform holds the job in Pending until a token becomes available.
How to diagnose:
Check with your license administrator to confirm whether the license pool is fully consumed.
Use the "Check availability" tool — If available for your license type, click Check availability in the job's Software Settings to see the current checkout status.
What you can do:
Wait for other jobs to release tokens. Your job starts automatically once a token is free.
If no other jobs appear to be using tokens and the job is still queued, contact Rescale Support — the license server configuration or Company License Proxy (CLP) may need review.
For background, see What is License Queuing? and Configuring License Feature Settings to Enable License Queuing on Rescale.
Platform or Provisioning Issue
Status Reason: "An error with the Rescale platform occurred" or "Error executing job" — or no status reason with an extended wait
In rare cases, a job may be stuck due to a platform-side provisioning defect or state synchronization problem.
Key signals:
The status reason mentions a platform error or job execution error
The job has been in Validating for more than 30 minutes with no status reason
Multiple jobs across different users in your organization are stuck simultaneously
What you can do:
Contact Rescale Support with the following:
Your Job ID
The status reason text (or note that none is shown)
How long the job has been stuck
Whether other users in your workspace are experiencing the same issue
Quick-Action Workaround — Duplicate and Resubmit
To unblock your work immediately while diagnosing the root cause:
Stop the stuck job (see How to Stop a Running Job on Rescale).
Duplicate the job using the Duplicate button on the job page.
Optionally, change the hardware type or region to avoid the same constraint.
Submit the new job.
This can resolve transient capacity or provisioning issues but does not fix underlying configuration problems such as core limits, license pool exhaustion, or quota caps. If the duplicated job also gets stuck, return to the diagnostic steps above.
When to Contact Rescale Support
Contact Rescale Support if:
Your job has been stuck for more than 2 hours and none of the above resolutions apply
The status reason mentions a platform error
Multiple jobs across your organization are stuck simultaneously
The issue requires a backend adjustment (quota increase, core limit change, or license server review)
When reaching out, share the job with Rescale Support (with permission to access) and include the Job ID, status reason text, how long the job has been stuck, and any steps already tried from this guide.