Skip to content
Introduction
Reference
Command line
dicer update

dicer update

Change a stopped instance’s definition

Synopsis

Changes the parts of a stopped instance’s definition that are given, and leaves the rest as it was. A list or map given – –env, –label, –publish, –mount – replaces the old one whole.

A command after – replaces the one the instance runs.

The restart policy alone can be changed while the instance runs: it applies the next time the instance ends.

dicer update NAME [-- COMMAND [ARG...]] [flags]

Examples 

  dicer update web --memory 2GiB --vcpus 2
  dicer update web --restart unless-stopped
  dicer update web -- /usr/sbin/nginx -g 'daemon off;'

Options 

  -i, --image string                   Container image reference
      --kernel string                  Kernel to boot with
      --kernel-args string             Kernel command line arguments
      --hypervisor-type string         Hypervisor: cloud-hypervisor or firecracker (default: cloud-hypervisor)
      --hypervisor-version string      Hypervisor version (default: the newest available)
      --vcpus int32                    Number of virtual CPUs
  -m, --memory string                  Memory, e.g. 512MiB or 2GiB
      --disk string                    Overlay disk size, e.g. 10GiB
      --network string                 Network to attach to
      --ip string                      Static IP address (default: assigned from the subnet)
  -p, --publish stringArray            Publish a guest port on the host, as [hostIP:]hostPort:guestPort[/tcp|udp] (repeatable)
      --mount stringArray              Mount a volume, host file or tmpfs, as [type=volume|file|tmpfs,][source=...,]target=/path[,readonly] (repeatable)
  -e, --env stringArray                Environment variable as KEY=VALUE, or KEY to pass this shell's value (repeatable)
      --env-file stringArray           Read environment variables from a file of KEY=VALUE lines (repeatable)
  -l, --label stringArray              Label as KEY=VALUE (repeatable)
      --hostname string                Guest hostname (default: the instance name)
      --restart string                 Restart policy when the instance ends on its own: no, on-failure[:max-retries], unless-stopped or always (default no)
      --rm                             Delete the instance once it stops, the daemon doing the deleting
      --health-cmd string              Command to check health with, run by /bin/sh in the guest
      --health-http string             Check health with an HTTP GET in the guest, as PORT[/path]; 2xx or 3xx is healthy
      --health-tcp int                 Check health by connecting to a TCP port in the guest
      --health-interval duration       Time between health checks (default 10s)
      --health-timeout duration        Time a health check may take (default 5s)
      --health-start-period duration   Time after a start in which failed checks do not count (default none)
      --health-retries int32           Failed checks in a row that make the instance unhealthy (default 3)
      --no-healthcheck                 Check no health, not even as the image says to
      --init-mode string               How the guest starts the command: auto, exec (as PID 1 of its own PID namespace) or systemd (default auto)
  -h, --help                           help for update

Options inherited from parent commands 

  -D, --debug              Trace every call to the daemon on stderr (or set $DICER_DEBUG)
  -r, --remote string      Daemon to talk to: a remote's name, or an address (unix:///PATH or HOST:PORT) (default $DICER_REMOTE, then the current remote, then local)
      --timeout duration   Give up on a call to the daemon after this long, e.g. 30s (0: never; streams such as logs -f and exec are not bounded)

SEE ALSO

  • dicer - Run virtual machines from container images
dicer stopdicer version