| Previous | Next |
| ERROR_INVALID_SHOWWIN_COMMAND | ERROR_NONPAGED_SYSTEM_RESOURCES |
ERROR_NO_SYSTEM_RESOURCES
A shared operating-system resource has been exhausted.
ERROR_NO_SYSTEM_RESOURCES is Win32 error 1450 (0x5AA). It is broader than a failed heap allocation. The unavailable resource may be system commit, kernel pool, page-table entries, desktop heap, handles, I/O bookkeeping, or another finite kernel-managed structure needed by the operation.
Signals that narrow the diagnosis
- the failure affects many unrelated processes at the same time
- restarting one leaking service temporarily restores the system
- handle, GDI, USER, pool, or commit counters rise steadily before failure
- the same workload succeeds after reboot but degrades over hours or days
- only a high-concurrency or large-I/O configuration reproduces the condition
What to capture before reboot
Collect total commit and commit limit, available memory, paged and nonpaged pool usage, system and per-process handle counts, USER/GDI object counts, process private bytes, and the failing API with request size. Preserve Resource Exhaustion Detector events and a time series from the period before the first failure. A post-reboot snapshot removes the most valuable evidence.
Diagnostic sequence
Determine whether the error is machine-wide or isolated to one process or session. Correlate the first occurrence with counters, deployments, driver activity, and workload changes. If pool growth is suspected, use supported kernel diagnostics such as pool tags and PoolMon; if handle growth is suspected, break counts down by process and object type.
Review request amplification as well. An accidental loop that creates thousands of windows, events, mappings, or outstanding I/O requests can exhaust a resource without a traditional memory leak. Validate quotas and concurrency limits before changing operating-system settings.
Recovery strategy
Stop or restart the leaking component when operationally safe, reduce workload fan-out, and reboot only when the exhausted resource cannot otherwise be reclaimed. The durable fix is to close the leaked object, bound queues, or correct the driver/application allocation path. Increasing a limit without understanding growth usually postpones the outage.
How this relates to codes 1451–1455
ERROR_NONPAGED_SYSTEM_RESOURCES, ERROR_PAGED_SYSTEM_RESOURCES, ERROR_WORKING_SET_QUOTA, ERROR_PAGEFILE_QUOTA, and ERROR_COMMITMENT_LIMIT identify more specific pressure domains. Error 1450 is the general result when the subsystem cannot provide the required shared resource.
Example
A service leaks event handles for every client reconnect. After several days, new process and file operations begin returning 1450 across the server. Handle-count telemetry identifies the service, and closing the per-session events during disconnect eliminates the system-wide failure.
References
- Microsoft: System Error Codes (1300–1699)
- Microsoft: Troubleshoot application or service memory leaks
- Microsoft: Process memory usage information
Looking for a different code? Search another status or error code.