From Linux Checkpoint / Restart Wiki
Revision as of 15:46, 17 August 2010 by Orenl (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

A feature being unsupported for checkpoint/restart could mean two things. It could mean that checkpoint on an application using the feature will fail. For instance, if an application has an open fd on an unsupported device or on a file of an unsupported fs, then checkpoint will fail with an informative error message. That is ok.

It could also mean that checkpoint will 'succeed', but restart will fail. Worse, it could mean that restart will 'succeed', but the application will misbehave after restart. For instance, we currently do not address mounts. So if an application has done 'mount(FILE1, FILE2, NULL, MS_BIND, 0)' before being checkpointed, then after restart FILE1 may no longer be bind-mounted onto FILE2.

Supported Architectures

  • x86-32, x86-64
  • PowerPC
  • s390x
  • ARM

Click here for information about architecture specific.

What is Implemented

Task and processess

  • multiple processes and threads
  • zombie processes
  • process ids (pid, pgid, sid)
  • futexes (pi, robust, plain)
  • signals (signal handling, pending signals)
  • credentials (userid, group, POSIX capabilities)


  • user-ns, ipc-ns, uts-ns, net-ns: multiple namespaces
  • pid-ns: a single namespace is supported (extension for nested pid-ns will be done in user-space).

File descriptors

  • regular files and directories:
    • ext2, ext3, ext4
    • affs, befs, bfs, btrfs, cramfs, ecryptfs, fat, freevxfs, hfs, hfsplus, hpfs, hppfs, iso, jffs2, jfs, regular, ntfs, openromfs, qnx4, reiserfs, romfs, squashfs, sysv, ubifs, udf, xfs
  • pipes, FIFOs
  • open devices:
    • /dev/null, zero, random, urandom
    • Unix98 ptys
  • special file descriptors:
    • epoll fd
    • event fd


  • unix sockets
  • ipv4, ipv6 sockets (except time-wait sockets)
  • veth, macvlan, loopback devices and addresses

IPv4 supports live migration between hosts.


    • message queues
    • semaphores (patches for 'undo' queued)
    • shared memory


  • Smack and SELinux LSM labels

What is NOT implemented

(please make each a link to a page describing issues, a design for solving, how we will detect it to refuse checkpoint if we will not support it, and "how ugly it will make the kernel". The table summarizes these details. )

Not Implemented Refuses to Checkpoint
Anticipated Solution Impact
mounts namespaces Yes Clean (the dirty work will be in mounts themselves)
mounts No postponed
time-wait inet sockets No

Clean, not high priority.

unlinked files and directories Yes

Not trivial, but shouldn't be ugly (discussion)

inotify Yes Probably gross due to watches attaching to inodes
fanotify Yes
dnotify Yes
FUSE Yes Clean but requires per-FUSE-filesystem support
network and distributed filesystems Yes
SYSV IPC: semaphore undo No Patches on the list already.
netlink Yes Clean, low priority.
new, expiremental, and/or
infrastructure-oriented network protocols
Yes Ignore
file locks Yes Patches on the list already
file leases Yes Patches on the list already
fowner+sigio Yes Trivial and clean.
ptraced tasks Yes Expect small impact on kernel code.
async-io Yes
restart of 64-bit task from
32-bit task and vice versa
Yes Probably in userspace
inet6 sockets Yes Previously supported, just need
to add it back in. minor.
Hardware devices
other than those mentioned above.
Includes mmap, /dev/foo, Infiniband devices
Yes (mostly) One idea, very low priority
System-specific files in sysfs
(things like UUID files)
time namespace N/A
Personal tools