// ************** Checkpoint/restart ***************
//
// This would limit lost work if the host is rebooted
// while a BUDA job is running.
// In other cases (e.g. the user exits the client)
// we pause the container and no work is lost.
//
// Podman and Docker have different interfaces:
// podman: https://podman.io/docs/checkpoint
//      podman container checkpoint <name>
//      podman container restore <name>
// docker: https://docs.docker.com/reference/cli/docker/checkpoint/
//      docker checkpoint create <cont_name> <checkpoint_name>
//      docker checkpoint ls   (lists checkpoints)
//      docker checkpoint rm   (delete checkpoint)
//      docker start --checkpoint <checkpoint_name> <cont_name>
//      (use <cont_name>_checkpoint as the checkpoint name)
//      need --security-opt=seccomp:unconfined in initial run command?
//
// In both cases the CRIU library is required.
//
// Note: I spent several days trying to get it to work on Podman
// with all sorts of variants like --pre-checkpoint, --keep etc.
// Whenever I did
// - create a checkpoint
// - compute some more
// - stop or kill the container
// - restore
// I always get
// Error: OCI runtime error: runc: criu failed: type RESTORE errno 0
//
// So I gave up; checkpoint/restart is only relevant to the reboot scenario,
// this is rare.
//
// I took out the checkpoint-related code 1/30/2026.
// If you want to see it, look at an older version

