Difference between revisions of "Integration"

From CRIU
Jump to navigation Jump to search
(Structure changes)
m (Systemd -> others)
Line 66: Line 66:
 
''See also: [[X applications]]
 
''See also: [[X applications]]
  
== Systemd ==
+
=== Systemd ===
 
{{Status|not started}}
 
{{Status|not started}}
 
Adrian suggested that
 
Adrian suggested that

Revision as of 17:59, 10 September 2016

CRIU is not so easy to be used as a standalone tool -- it works best integrated into other software. This page lists such software and provides details about the current status.

Container platforms

Virtuozzo/OpenVZ

Status: ready

Currently, vzctl supports CRIU for checkpoint/restore of upstream containers (i.e. when non-OpenVZ kernel is used). Commands vzctl suspend and vzctl restore fully work. Live migration doesn't work yet as it requires support for vzctl suspend {--suspend, --dump, --kill, --restore} which is not yet implemented as it requires a separate daemon to hold the state of a partially checkpointed container (or an ability from criu tool to do that).

LXC/LXD

Status: ready

Docker

Status: in progress

CoreOS Rocket

Status: not started

Tools and utilities

The file utility

Status: ready

  • Starting from v1.6 new images (v1.1) will be generated
  • File utility starting from 5.23 will support these

Screen/TMUX

Status: stalled

  • Jerome did this some time ago

Shell

Status: not started

It would be nice to have bash (or other shell) to launch criu with --restore-sibling option and get new kid processes from it.

Other

OpenMPI

Status: stalled

  • Adrian Reber did first version of patches

Subgraph OS

Status: not started

  • Subgraph OS is a desktop operation system uses containers for users applications.

SLURM workload manager

Status: ready

Wayland/Weston

Status: stalled

  • Ruslan Kuprieiev plans to patch Weston to let CRIU C/R graphical apps

See also: X applications

Systemd

Status: not started Adrian suggested that

Migrating
processes from one system to another works, depending on the process,
pretty good. Migrating a process under systemd's control might be
possible by just killing the process on the source side but it cannot
become a child process of systemd on the destination of the migration
without systemd knowing how to restore a process and thus making it a
child process of systemd (--restore-sibling).