Not expecting a fix on your end, but sharing this in case others hit the same thing.
Environment
Npgsql version: 10.0.3
Runtime: .NET 10.0.8, Alpine 3.23
GssEncryptionMode: default (not set)
Observed behavior
After upgrading to 10.0.3, our Kubernetes pods restart intermittently — roughly once per hour — due to a SIGSEGV. The crashes surface first as failed health checks before the pod is terminated.
Root cause (suspected)
We believe this is caused by a bug in krb5 1.22.1 on Alpine 3.23, not in Npgsql itself. The relevant upstream issues are:
The reason this appeared after the Npgsql upgrade is possibly the changes to the GSS encryption in 10.0.3.
Mitigation
Setting GssEncryptionMode=Disable in the connection string resolves the crashes entirely.
Not expecting a fix on your end, but sharing this in case others hit the same thing.
Environment
Npgsql version: 10.0.3
Runtime: .NET 10.0.8, Alpine 3.23
GssEncryptionMode: default (not set)
Observed behavior
After upgrading to 10.0.3, our Kubernetes pods restart intermittently — roughly once per hour — due to a SIGSEGV. The crashes surface first as failed health checks before the pod is terminated.
Root cause (suspected)
We believe this is caused by a bug in krb5 1.22.1 on Alpine 3.23, not in Npgsql itself. The relevant upstream issues are:
The reason this appeared after the Npgsql upgrade is possibly the changes to the GSS encryption in 10.0.3.
Mitigation
Setting
GssEncryptionMode=Disablein the connection string resolves the crashes entirely.