Understand the failure.
Verify the repair.
Small, reproducible software failures. The broken implementation, the fix that didn’t work, and the one that passed—preserved together.
Explore the cases ↓How results are verified ↗WHAT THE ARCHIVE CONTAINS
100840 executable cases. 20168 are open.
Every case records the implementation that fails, the fix that did not work, and the repair that passed its checks—with recorded outputs and source hashes. This release adds 100840 cases across 20168 failure mechanisms and 254 domains.
The open tier gives you the failure and the unsuccessful fix for one case in every mechanism. The remaining 80672 cases, 5 variants per mechanism, are member-only: the verified repair, its recorded checks, and the full fixture suite are held in the member archive. Read the methodology ↗
A RECORD OF WHAT WENT WRONG
Browse the archive / 100840
Enter a dependency bulkhead: Concurrent calls exceed the dependency execution limit · case 01
The bulkhead admission operation is admitted even though concurrent calls exceed the dependency execution limit.
Enter a dependency bulkhead: Concurrent calls exceed the dependency execution limit · case 02
The bulkhead admission operation is admitted even though concurrent calls exceed the dependency execution limit.
Enter a dependency bulkhead: Concurrent calls exceed the dependency execution limit · case 03
The bulkhead admission operation is admitted even though concurrent calls exceed the dependency execution limit.
Enter a dependency bulkhead: Concurrent calls exceed the dependency execution limit · case 04
The bulkhead admission operation is admitted even though concurrent calls exceed the dependency execution limit.
Enter a dependency bulkhead: Concurrent calls exceed the dependency execution limit · case 05
The bulkhead admission operation is admitted even though concurrent calls exceed the dependency execution limit.
Enter a dependency bulkhead: Waiting callers exceed the bounded waiting queue · case 01
The bulkhead admission operation is admitted even though waiting callers exceed the bounded waiting queue.
Enter a dependency bulkhead: Waiting callers exceed the bounded waiting queue · case 02
The bulkhead admission operation is admitted even though waiting callers exceed the bounded waiting queue.
Enter a dependency bulkhead: Waiting callers exceed the bounded waiting queue · case 03
The bulkhead admission operation is admitted even though waiting callers exceed the bounded waiting queue.
Enter a dependency bulkhead: Waiting callers exceed the bounded waiting queue · case 04
The bulkhead admission operation is admitted even though waiting callers exceed the bounded waiting queue.
Enter a dependency bulkhead: Waiting callers exceed the bounded waiting queue · case 05
The bulkhead admission operation is admitted even though waiting callers exceed the bounded waiting queue.
Enter a dependency bulkhead: A waiting call starts after its queue deadline · case 01
The bulkhead admission operation is admitted even though a waiting call starts after its queue deadline.
Enter a dependency bulkhead: A waiting call starts after its queue deadline · case 02
The bulkhead admission operation is admitted even though a waiting call starts after its queue deadline.
Enter a dependency bulkhead: A waiting call starts after its queue deadline · case 03
The bulkhead admission operation is admitted even though a waiting call starts after its queue deadline.
Enter a dependency bulkhead: A waiting call starts after its queue deadline · case 04
The bulkhead admission operation is admitted even though a waiting call starts after its queue deadline.
Enter a dependency bulkhead: A waiting call starts after its queue deadline · case 05
The bulkhead admission operation is admitted even though a waiting call starts after its queue deadline.
Enter a dependency bulkhead: Normal traffic consumes capacity reserved for recovery · case 01
The bulkhead admission operation is admitted even though normal traffic consumes capacity reserved for recovery.
Enter a dependency bulkhead: Normal traffic consumes capacity reserved for recovery · case 02
The bulkhead admission operation is admitted even though normal traffic consumes capacity reserved for recovery.
Enter a dependency bulkhead: Normal traffic consumes capacity reserved for recovery · case 03
The bulkhead admission operation is admitted even though normal traffic consumes capacity reserved for recovery.
Enter a dependency bulkhead: Normal traffic consumes capacity reserved for recovery · case 04
The bulkhead admission operation is admitted even though normal traffic consumes capacity reserved for recovery.
Enter a dependency bulkhead: Normal traffic consumes capacity reserved for recovery · case 05
The bulkhead admission operation is admitted even though normal traffic consumes capacity reserved for recovery.
Enter a dependency bulkhead: Unrelated dependencies share a bulkhead accounting key · case 01
The bulkhead admission operation is admitted even though unrelated dependencies share a bulkhead accounting key.
Enter a dependency bulkhead: Unrelated dependencies share a bulkhead accounting key · case 02
The bulkhead admission operation is admitted even though unrelated dependencies share a bulkhead accounting key.
Enter a dependency bulkhead: Unrelated dependencies share a bulkhead accounting key · case 03
The bulkhead admission operation is admitted even though unrelated dependencies share a bulkhead accounting key.
Enter a dependency bulkhead: Unrelated dependencies share a bulkhead accounting key · case 04
The bulkhead admission operation is admitted even though unrelated dependencies share a bulkhead accounting key.
Enter a dependency bulkhead: Unrelated dependencies share a bulkhead accounting key · case 05
The bulkhead admission operation is admitted even though unrelated dependencies share a bulkhead accounting key.
Evict an unhealthy backend: One intermittent failure causes premature eviction · case 01
The health eviction operation is admitted even though one intermittent failure causes premature eviction.
Evict an unhealthy backend: One intermittent failure causes premature eviction · case 02
The health eviction operation is admitted even though one intermittent failure causes premature eviction.
Evict an unhealthy backend: One intermittent failure causes premature eviction · case 03
The health eviction operation is admitted even though one intermittent failure causes premature eviction.
Evict an unhealthy backend: One intermittent failure causes premature eviction · case 04
The health eviction operation is admitted even though one intermittent failure causes premature eviction.
Evict an unhealthy backend: One intermittent failure causes premature eviction · case 05
The health eviction operation is admitted even though one intermittent failure causes premature eviction.
Evict an unhealthy backend: Eviction removes the last required healthy capacity · case 01
The health eviction operation is admitted even though eviction removes the last required healthy capacity.
Evict an unhealthy backend: Eviction removes the last required healthy capacity · case 02
The health eviction operation is admitted even though eviction removes the last required healthy capacity.
Evict an unhealthy backend: Eviction removes the last required healthy capacity · case 03
The health eviction operation is admitted even though eviction removes the last required healthy capacity.
Evict an unhealthy backend: Eviction removes the last required healthy capacity · case 04
The health eviction operation is admitted even though eviction removes the last required healthy capacity.
Evict an unhealthy backend: Eviction removes the last required healthy capacity · case 05
The health eviction operation is admitted even though eviction removes the last required healthy capacity.
Evict an unhealthy backend: Old failure samples evict a recovered backend · case 01
The health eviction operation is admitted even though old failure samples evict a recovered backend.
Evict an unhealthy backend: Old failure samples evict a recovered backend · case 02
The health eviction operation is admitted even though old failure samples evict a recovered backend.
Evict an unhealthy backend: Old failure samples evict a recovered backend · case 03
The health eviction operation is admitted even though old failure samples evict a recovered backend.
Evict an unhealthy backend: Old failure samples evict a recovered backend · case 04
The health eviction operation is admitted even though old failure samples evict a recovered backend.
Evict an unhealthy backend: Old failure samples evict a recovered backend · case 05
The health eviction operation is admitted even though old failure samples evict a recovered backend.
Evict an unhealthy backend: Correlated failures evict more than the allowed pool fraction · case 01
The health eviction operation is admitted even though correlated failures evict more than the allowed pool fraction.
Evict an unhealthy backend: Correlated failures evict more than the allowed pool fraction · case 02
The health eviction operation is admitted even though correlated failures evict more than the allowed pool fraction.
Evict an unhealthy backend: Correlated failures evict more than the allowed pool fraction · case 03
The health eviction operation is admitted even though correlated failures evict more than the allowed pool fraction.
Evict an unhealthy backend: Correlated failures evict more than the allowed pool fraction · case 04
The health eviction operation is admitted even though correlated failures evict more than the allowed pool fraction.
Evict an unhealthy backend: Correlated failures evict more than the allowed pool fraction · case 05
The health eviction operation is admitted even though correlated failures evict more than the allowed pool fraction.
Evict an unhealthy backend: A failed probe evicts a replacement process at the same address · case 01
The health eviction operation is admitted even though a failed probe evicts a replacement process at the same address.
Evict an unhealthy backend: A failed probe evicts a replacement process at the same address · case 02
The health eviction operation is admitted even though a failed probe evicts a replacement process at the same address.
Evict an unhealthy backend: A failed probe evicts a replacement process at the same address · case 03
The health eviction operation is admitted even though a failed probe evicts a replacement process at the same address.
Evict an unhealthy backend: A failed probe evicts a replacement process at the same address · case 04
The health eviction operation is admitted even though a failed probe evicts a replacement process at the same address.
Evict an unhealthy backend: A failed probe evicts a replacement process at the same address · case 05
The health eviction operation is admitted even though a failed probe evicts a replacement process at the same address.
Finish draining a service instance: Shutdown still permits new work to enter · case 01
The graceful drain operation is admitted even though shutdown still permits new work to enter.
Finish draining a service instance: Shutdown still permits new work to enter · case 02
The graceful drain operation is admitted even though shutdown still permits new work to enter.
Finish draining a service instance: Shutdown still permits new work to enter · case 03
The graceful drain operation is admitted even though shutdown still permits new work to enter.
Finish draining a service instance: Shutdown still permits new work to enter · case 04
The graceful drain operation is admitted even though shutdown still permits new work to enter.
Finish draining a service instance: Shutdown still permits new work to enter · case 05
The graceful drain operation is admitted even though shutdown still permits new work to enter.
Finish draining a service instance: Shutdown terminates in-flight requests prematurely · case 01
The graceful drain operation is admitted even though shutdown terminates in-flight requests prematurely.
Finish draining a service instance: Shutdown terminates in-flight requests prematurely · case 02
The graceful drain operation is admitted even though shutdown terminates in-flight requests prematurely.
Finish draining a service instance: Shutdown terminates in-flight requests prematurely · case 03
The graceful drain operation is admitted even though shutdown terminates in-flight requests prematurely.
Finish draining a service instance: Shutdown terminates in-flight requests prematurely · case 04
The graceful drain operation is admitted even though shutdown terminates in-flight requests prematurely.
Finish draining a service instance: Shutdown terminates in-flight requests prematurely · case 05
The graceful drain operation is admitted even though shutdown terminates in-flight requests prematurely.
Finish draining a service instance: Shutdown completes while routing still advertises the instance · case 01
The graceful drain operation is admitted even though shutdown completes while routing still advertises the instance.
Finish draining a service instance: Shutdown completes while routing still advertises the instance · case 02
The graceful drain operation is admitted even though shutdown completes while routing still advertises the instance.
Finish draining a service instance: Shutdown completes while routing still advertises the instance · case 03
The graceful drain operation is admitted even though shutdown completes while routing still advertises the instance.
Finish draining a service instance: Shutdown completes while routing still advertises the instance · case 04
The graceful drain operation is admitted even though shutdown completes while routing still advertises the instance.
Finish draining a service instance: Shutdown completes while routing still advertises the instance · case 05
The graceful drain operation is admitted even though shutdown completes while routing still advertises the instance.
Finish draining a service instance: Shutdown discards buffered durable work · case 01
The graceful drain operation is admitted even though shutdown discards buffered durable work.
Finish draining a service instance: Shutdown discards buffered durable work · case 02
The graceful drain operation is admitted even though shutdown discards buffered durable work.
Finish draining a service instance: Shutdown discards buffered durable work · case 03
The graceful drain operation is admitted even though shutdown discards buffered durable work.
Finish draining a service instance: Shutdown discards buffered durable work · case 04
The graceful drain operation is admitted even though shutdown discards buffered durable work.
Finish draining a service instance: Shutdown discards buffered durable work · case 05
The graceful drain operation is admitted even though shutdown discards buffered durable work.
Finish draining a service instance: Shutdown abandons child workers holding resources · case 01
The graceful drain operation is admitted even though shutdown abandons child workers holding resources.
Finish draining a service instance: Shutdown abandons child workers holding resources · case 02
The graceful drain operation is admitted even though shutdown abandons child workers holding resources.
Finish draining a service instance: Shutdown abandons child workers holding resources · case 03
The graceful drain operation is admitted even though shutdown abandons child workers holding resources.
Finish draining a service instance: Shutdown abandons child workers holding resources · case 04
The graceful drain operation is admitted even though shutdown abandons child workers holding resources.
Finish draining a service instance: Shutdown abandons child workers holding resources · case 05
The graceful drain operation is admitted even though shutdown abandons child workers holding resources.
Promote a standby after primary failure: Promotion permits two writable primaries · case 01
The failover promote operation is admitted even though promotion permits two writable primaries.
Promote a standby after primary failure: Promotion permits two writable primaries · case 02
The failover promote operation is admitted even though promotion permits two writable primaries.
Promote a standby after primary failure: Promotion permits two writable primaries · case 03
The failover promote operation is admitted even though promotion permits two writable primaries.
Promote a standby after primary failure: Promotion permits two writable primaries · case 04
The failover promote operation is admitted even though promotion permits two writable primaries.
Promote a standby after primary failure: Promotion permits two writable primaries · case 05
The failover promote operation is admitted even though promotion permits two writable primaries.
Promote a standby after primary failure: Standby lag exceeds the declared data-loss allowance · case 01
The failover promote operation is admitted even though standby lag exceeds the declared data-loss allowance.
Promote a standby after primary failure: Standby lag exceeds the declared data-loss allowance · case 02
The failover promote operation is admitted even though standby lag exceeds the declared data-loss allowance.
Promote a standby after primary failure: Standby lag exceeds the declared data-loss allowance · case 03
The failover promote operation is admitted even though standby lag exceeds the declared data-loss allowance.
Promote a standby after primary failure: Standby lag exceeds the declared data-loss allowance · case 04
The failover promote operation is admitted even though standby lag exceeds the declared data-loss allowance.
Promote a standby after primary failure: Standby lag exceeds the declared data-loss allowance · case 05
The failover promote operation is admitted even though standby lag exceeds the declared data-loss allowance.
Promote a standby after primary failure: A read-only standby is advertised as write primary · case 01
The failover promote operation is admitted even though a read-only standby is advertised as write primary.
Promote a standby after primary failure: A read-only standby is advertised as write primary · case 02
The failover promote operation is admitted even though a read-only standby is advertised as write primary.
Promote a standby after primary failure: A read-only standby is advertised as write primary · case 03
The failover promote operation is admitted even though a read-only standby is advertised as write primary.
Promote a standby after primary failure: A read-only standby is advertised as write primary · case 04
The failover promote operation is admitted even though a read-only standby is advertised as write primary.
Promote a standby after primary failure: A read-only standby is advertised as write primary · case 05
The failover promote operation is admitted even though a read-only standby is advertised as write primary.
Promote a standby after primary failure: A delayed promotion reverses a newer failover · case 01
The failover promote operation is admitted even though a delayed promotion reverses a newer failover.
Promote a standby after primary failure: A delayed promotion reverses a newer failover · case 02
The failover promote operation is admitted even though a delayed promotion reverses a newer failover.
Promote a standby after primary failure: A delayed promotion reverses a newer failover · case 03
The failover promote operation is admitted even though a delayed promotion reverses a newer failover.
Promote a standby after primary failure: A delayed promotion reverses a newer failover · case 04
The failover promote operation is admitted even though a delayed promotion reverses a newer failover.
Promote a standby after primary failure: A delayed promotion reverses a newer failover · case 05
The failover promote operation is admitted even though a delayed promotion reverses a newer failover.
Promote a standby after primary failure: Promotion succeeds while writes still route to the old primary · case 01
The failover promote operation is admitted even though promotion succeeds while writes still route to the old primary.
Promote a standby after primary failure: Promotion succeeds while writes still route to the old primary · case 02
The failover promote operation is admitted even though promotion succeeds while writes still route to the old primary.
Promote a standby after primary failure: Promotion succeeds while writes still route to the old primary · case 03
The failover promote operation is admitted even though promotion succeeds while writes still route to the old primary.
Promote a standby after primary failure: Promotion succeeds while writes still route to the old primary · case 04
The failover promote operation is admitted even though promotion succeeds while writes still route to the old primary.
Promote a standby after primary failure: Promotion succeeds while writes still route to the old primary · case 05
The failover promote operation is admitted even though promotion succeeds while writes still route to the old primary.
INSPECTABLE BY DESIGN
Every result has a runnable source.
Runnable implementations with recorded outputs, source hashes, and explicit contracts. Related variants share a failure mechanism and belong together in evaluation splits.
Read the methodology ↗