Covers application checkpoint/restart, overall design, interfaces,
usage, shared objects, and and checkpoint image format.
Changelog[v19-rc1]:
- Update documentation and examples for new syscalls API
- [Liu Alexander] Fix typos
- [Serge Hallyn] Update checkpoint image format
Changelog[v16]:
- Update documentation
- Unify into readme.txt and usage.txt
Changelog[v14]:
- Discard the 'h.parent' field
- New image format (shared objects appear before they are referenced
unless they are compound)
Changelog[v8]:
- Split into multiple files in Documentation/checkpoint/...
- Extend documentation, fix typos and comments from feedback
Signed-off-by: Oren Laadan <orenl@xxxxxxxxxxxxxxx>
Signed-off-by: Dave Hansen <dave@xxxxxxxxxxxxxxxxxx>
Acked-by: Serge E. Hallyn <serue@xxxxxxxxxx>
Tested-by: Serge E. Hallyn <serue@xxxxxxxxxx>
---
Documentation/checkpoint/checkpoint.c | 38 +++
Documentation/checkpoint/readme.txt | 370 ++++++++++++++++++++++++++++
Documentation/checkpoint/self_checkpoint.c | 69 +++++
Documentation/checkpoint/self_restart.c | 40 +++
Documentation/checkpoint/usage.txt | 247 +++++++++++++++++++
5 files changed, 764 insertions(+), 0 deletions(-)
create mode 100644 Documentation/checkpoint/checkpoint.c
create mode 100644 Documentation/checkpoint/readme.txt
create mode 100644 Documentation/checkpoint/self_checkpoint.c
create mode 100644 Documentation/checkpoint/self_restart.c
create mode 100644 Documentation/checkpoint/usage.txt
diff --git a/Documentation/checkpoint/checkpoint.c b/Documentation/checkpoint/checkpoint.c
new file mode 100644
index 0000000..8560f30
--- /dev/null
+++ b/Documentation/checkpoint/checkpoint.c
@@ -0,0 +1,38 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/syscall.h>
+
+#include <linux/checkpoint.h>
+
+static inline int checkpoint(pid_t pid, int fd, unsigned long flags)
+{
+ return syscall(__NR_checkpoint, pid, fd, flags);
+}
+
+int main(int argc, char *argv[])
+{
+ pid_t pid;
+ int ret;
+
+ if (argc != 2) {
+ printf("usage: ckpt PID\n");
+ exit(1);
+ }
+
+ pid = atoi(argv[1]);
+ if (pid <= 0) {
+ printf("invalid pid\n");
+ exit(1);
+ }
+
+ ret = checkpoint(pid, STDOUT_FILENO, CHECKPOINT_SUBTREE);
+
+ if (ret < 0)
+ perror("checkpoint");
+ else
+ printf("checkpoint id %d\n", ret);
+
+ return (ret > 0 ? 0 : 1);
+}
diff --git a/Documentation/checkpoint/readme.txt b/Documentation/checkpoint/readme.txt
new file mode 100644
index 0000000..4fa5560
--- /dev/null
+++ b/Documentation/checkpoint/readme.txt
@@ -0,0 +1,370 @@
+
+ Checkpoint-Restart support in the Linux kernel
+ ==========================================================
+
+Copyright (C) 2008-2010 Oren Laadan
+
+Author: Oren Laadan <orenl@xxxxxxxxxxxxxxx>
+
+License: The GNU Free Documentation License, Version 1.2
+ (dual licensed under the GPL v2)
+
+Contributors: Oren Laadan <orenl@xxxxxxxxxxxxxxx>
+ Serge Hallyn <serue@xxxxxxxxxx>
+ Dan Smith <danms@xxxxxxxxxx>
+ Matt Helsley <matthltc@xxxxxxxxxx>
+ Nathan Lynch <ntl@xxxxxxxxx>
+ Sukadev Bhattiprolu <sukadev@xxxxxxxxxxxxxxxxxx>
+ Dave Hansen <dave@xxxxxxxxxxxxxxxxxx>
+
+
+Introduction
+============
+
+Application checkpoint/restart [C/R] is the ability to save the state
+of a running application so that it can later resume its execution
+from the time at which it was checkpointed. An application can be
+migrated by checkpointing it on one machine and restarting it on
+another. C/R can provide many potential benefits:
+
+* Failure recovery: by rolling back to a previous checkpoint
+
+* Improved response time: by restarting applications from checkpoints
+ instead of from scratch.
+
+* Improved system utilization: by suspending long running CPU
+ intensive jobs and resuming them when load decreases.
+
+* Fault resilience: by migrating applications off faulty hosts.
+
+* Dynamic load balancing: by migrating applications to less loaded
+ hosts.
+
+* Improved service availability and administration: by migrating
+ applications before host maintenance so that they continue to run
+ with minimal downtime
+
+* Time-travel: by taking periodic checkpoints and restarting from
+ any previous checkpoint.
+
+Compared to hypervisor approaches, application C/R is more lightweight
+since it need only save the state associated with applications, while
+operating system data structures (e.g. buffer cache, drivers state
+and the like) are uninteresting.
+
+
+Overall design
+==============
+
+Checkpoint and restart are done in the kernel as much as possible.
+Two new system calls are introduced to provide C/R: sys_checkpoint()
+and sys_restart(). They both operate on a process tree (hierarchy),
+either a whole container or a subtree of a container.
+
+Checkpointing entire containers ensures that there are no dependencies
+on anything outside the container, which guarantees that a matching
+restart will succeed (assuming that the file system state remains
+consistent). However, it requires that users will always run the tasks
+that they wish to checkpoint inside containers. This is ideal for,
+e.g., private virtual servers and the like.
+
+In contrast, when checkpointing a subtree of a container it is up to
+the user to ensure that dependencies either don't exist or can be
+safely ignored. This is useful, for instance, for HPC scenarios or
+even a user that would like to periodically checkpoint a long-running
+batch job.
+
+An additional system call, a la madvise(), is planned, so that tasks
+can advise the kernel how to handle specific resources. For instance,
+a task could ask to skip a memory area at checkpoint to save space,
+or to use a preset file descriptor at restart instead of restoring it
+from the checkpoint image. It will provide the flexibility that is
+particularly useful to address the needs of a diverse crowd of users
+and use-cases.
+
+Syscall sys_checkpoint() is given a pid that indicates the top of the
+hierarchy, a file descriptor to store the image, and flags. The code
+serializes internal user- and kernel-state and writes it out to the
+file descriptor. The resulting image is stream-able. The processes are
+expected to be frozen for the duration of the checkpoint.
+
+In general, a checkpoint consists of 5 steps:
+1. Pre-dump
+2. Freeze the container/subtree
+3. Save tasks' and kernel state <-- sys_checkpoint()
+4. Thaw (or kill) the container/subtree
+5. Post-dump
+
+Step 3 is done by calling sys_checkpoint(). Steps 1 and 5 are an
+optimization to reduce application downtime. In particular, "pre-dump"
+works before freezing the container, e.g. the pre-copy for live
+migration, and "post-dump" works after the container resumes
+execution, e.g. write-back the data to secondary storage.
+
+The kernel exports a relatively opaque 'blob' of data to userspace
+which can then be handed to the new kernel at restart time. The
+'blob' contains data and state of select portions of kernel structures
+such as VMAs and mm_structs, as well as copies of the actual memory
+that the tasks use. Any changes in this blob's format between kernel
+revisions can be handled by an in-userspace conversion program.
+
+To restart, userspace first create a process hierarchy that matches
+that of the checkpoint, and each task calls sys_restart(). The syscall
+reads the saved kernel state from a file descriptor, and re-creates
+the resources that the tasks need to resume execution. The restart
+code is executed by each task that is restored in the new hierarchy to
+reconstruct its own state.
+
+In general, a restart consists of 3 steps:
+1. Create hierarchy
+2. Restore tasks' and kernel state <-- sys_restart()
+3. Resume userspace (or freeze tasks)
+
+Because the process hierarchy, during restart in created in userspace,
+the restarting tasks have the flexibility to prepare before calling
+sys_restart().
+
+
+Checkpoint image format
+=======================
+
+The checkpoint image format is built of records that consist of a
+pre-header identifying its contents, followed by a payload. This
+format allow userspace tools to easily parse and skip through the
+image without requiring intimate knowledge of the data. It will also
+be handy to enable parallel checkpointing in the future where multiple
+threads interleave data from multiple processes into a single stream.
+
+The pre-header is defined by 'struct ckpt_hdr' as follows: @type
+identifies the type of the payload, @len tells its length in bytes
+including the pre-header.
+
+struct ckpt_hdr {
+ __s32 type;
+ __s32 len;
+};
+
+The pre-header must be the first component in all other headers. For
+instance, the task data is saved in 'struct ckpt_hdr_task', which
+looks something like this:
+
+struct ckpt_hdr_task {
+ struct ckpt_hdr h;
+ __u32 pid;
+ ...
+};
+
+THE IMAGE FORMAT IS EXPECTED TO CHANGE over time as more features are
+supported, or as existing features change in the kernel and require to
+adjust their representation. Any such changes will be be handled by
+in-userspace conversion tools.
+
+The general format of the checkpoint image is as follows:
+* Image header
+* Container configuration
+* Task hierarchy
+* Tasks' state
+* Image trailer
+
+The image always begins with a general header that holds a magic
+number, an architecture identifier (little endian format), a format
+version number (@rev), followed by information about the kernel
+(currently version and UTS data). It also holds the time of the
+checkpoint and the flags given to sys_checkpoint(). This header is
+followed by an arch-specific header.
+
+The container configuration section containers information that is
+global to the container. Security (LSM) configuration is one example.
+Network configuration and container-wide mounts may also go here, so
+that the userspace restart coordinator can re-create a suitable
+environment.
+
+The task hierarchy comes next so that userspace tools can read it
+early (even from a stream) and re-create the restarting tasks. This is
+basically an array of all checkpointed tasks, and their relationships
+(parent, siblings, threads, etc).
+
+Then the state of all tasks is saved, in the order that they appear in
+the tasks array above. For each state, we save data like task_struct,
+namespaces, open files, memory layout, memory contents, cpu state,
+signals and signal handlers, etc. For resources that are shared among
+multiple processes, we first checkpoint said resource (and only once),
+and in the task data we give a reference to it. More about shared
+resources below.
+
+Finally, the image always ends with a trailer that holds a (different)
+magic number, serving for sanity check.
+
+
+Shared objects
+==============
+
+Many resources may be shared by multiple tasks (e.g. file descriptors,
+memory address space, etc), or even have multiple references from
+other resources (e.g. a single inode that represents two ends of a
+pipe).
+
+Shared objects are tracked using a hash table (objhash) to ensure that
+they are only checkpointed or restored once. To handle a shared
+object, it is first looked up in the hash table, to determine if is
+the first encounter or a recurring appearance. The hash table itself
+is not saved as part of the checkpoint image: it is constructed
+dynamically during both checkpoint and restart, and discarded at the
+end of the operation.
+
+During checkpoint, when a shared object is encountered for the first
+time, it is inserted to the hash table, indexed by its kernel address.
+It is assigned an identifier (@objref) in order of appearance, and
+then its state is saved. Subsequent lookups of that object in the hash
+will yield that entry, in which case only the @objref is saved, as
+opposed the entire state of the object.
+
+During restart, shared objects are indexed by their @objref as given
+during the checkpoint. On the first appearance of each shared object,
+a new resource will be created and its state restored from the image.
+Then the object is added to the hash table. Subsequent lookups of the
+same unique identifier in the hash table will yield that entry, and
+then the existing object instance is reused instead of creating
+a new one.
+
+The hash grabs a reference to each object that is inserted, and
+maintains this reference for the entire lifetime of the hash. Thus,
+it is always safe to reference an object that is stored in the hash.
+The hash is "one-way" in the sense that objects that are added are
+never deleted from the hash until the hash is discarded. This, in
+turn, happens only when the checkpoint (or restart) terminates.
+
+Shared objects are thus saved when they are first seen, and _before_
+the parent object that uses them. Therefore by the time the parent
+objects needs them, they should already be in the objhash. The one
+exception is when more than a single shared resource will be restarted
+at once (e.g. like the two ends of a pipe, or all the namespaces in an
+nsproxy). In this case the parent object is dumped first followed by
+the individual sub-resources).
+
+The checkpoint image is stream-able, meaning that restarting from it
+may not require lseek(). This is enforced at checkpoint time, by
+carefully selecting the order of shared objects, to respect the rule
+that an object is always saved before the objects that refers to it.
+
+
+Memory contents format
+======================
+
+The memory contents of a given memory address space (->mm) is dumped
+as a sequence of vma objects, represented by 'struct ckpt_hdr_vma'.
+This header details the vma properties, and a reference to a file
+(if file backed) or an inode (or shared memory) object.
+
+The vma header is followed by the actual contents - but only those
+pages that need to be saved, i.e. dirty pages. They are written in
+chunks of data, where each chunks contains a header that indicates
+that number of pages in the chunk, followed by an array of virtual
+addresses and then an array of actual page contents. The last chunk
+holds zero pages.
+
+To illustrate this, consider a single simple task with two vmas: one
+is file mapped with two dumped pages, and the other is anonymous with
+three dumped pages. The memory dump will look like this:
+
+ ckpt_hdr + ckpt_hdr_vma
+ ckpt_hdr_pgarr (nr_pages = 2)
+ addr1, addr2
+ page1, page2
+ ckpt_hdr_pgarr (nr_pages = 0)
+ ckpt_hdr + ckpt_hdr_vma
+ ckpt_hdr_pgarr (nr_pages = 3)
+ addr3, addr4, addr5
+ page3, page4, page5
+ ckpt_hdr_pgarr (nr_pages = 0)
+
+
+Error handling
+==============
+
+Both checkpoint and restart operations may fail due to a variety of
+reasons. Using a simple, single return value from the system call is
+insufficient to report the reason of a failure.
+
+Instead, both sys_checkpoint() and sys_restart() accept an additional
+argument - a file descriptor to which the kernel writes diagnostic
+and debugging information. Both the checkpoint and restart userspace
+utilities have options to specify a filename to store this log.
+
+In addition, checkpoint provides informative status report upon
+failure in the checkpoint image in the form of (one or more) error
+objects, 'struct ckpt_hdr_err'. An error objects consists of a
+mandatory pre-header followed by a null character ('\0'), and then a
+string that describes the error. By default, if an error occurs, this
+will be the last object written to the checkpoint image.
+
+Upon failure, the caller can examine the image (e.g. with 'ckptinfo')
+and extract the detailed error message. The leading '\0' is useful if
+one wants to seek back from the end of the checkpoint image, instead
+of parsing the entire image separately.
+
+
+Security
+========
+
+The main question is whether sys_checkpoint() and sys_restart()
+require privileged or unprivileged operation.
+
+Early versions checked capable(CAP_SYS_ADMIN) assuming that we would
+attempt to remove the need for privilege, so that all users could
+safely use it. Arnd Bergmann pointed out that it'd make more sense to
+let unprivileged users use them now, so that we'll be more careful
+about the security as patches roll in.
+
+Checkpoint: the main concern is whether a task that performs the
+checkpoint of another task has sufficient privileges to access its
+state. We address this by requiring that the checkpointer task will be
+able to ptrace the target task, by means of ptrace_may_access() with
+access mode.
+
+Restart: the main concern is that we may allow an unprivileged user to
+feed the kernel with random data. To this end, the restart works in a
+way that does not skip the usual security checks. Task credentials,
+i.e. euid, reuid, and LSM security contexts currently come from the
+caller, not the checkpoint image. As credentials are restored too,
+the ability of a task that calls sys_restore() to setresuid/setresgid
+to those values must be checked.
+
+Keeping the restart procedure to operate within the limits of the
+caller's credentials means that there various scenarios that cannot
+be supported. For instance, a setuid program that opened a protected
+log file and then dropped privileges will fail the restart, because
+the user won't have enough credentials to reopen the file. In these
+cases, we should probably treat restarting like inserting a kernel
+module: surely the user can cause havoc by providing incorrect data,
+but then again we must trust the root account.
+
+So that's why we don't want CAP_SYS_ADMIN required up-front. That way
+we will be forced to more carefully review each of those features.
+However, this can be controlled with a sysctl-variable.
+
+
+Kernel interfaces
+=================
+
+* To checkpoint a vma, the 'struct vm_operations_struct' needs to
+ provide a method ->checkpoint:
+ int checkpoint(struct ckpt_ctx *, struct vma_struct *)
+ Restart requires a matching (exported) restore:
+ int restore(struct ckpt_ctx *, struct mm_struct *, struct ckpt_hdr_vma *)
+
+* To checkpoint a file, the 'struct file_operations' needs to provide
+ the methods ->checkpoint and ->collect:
+ int checkpoint(struct ckpt_ctx *, struct file *)
+ int collect(struct ckpt_ctx *, struct file *)
+ Restart requires a matching (exported) restore:
+ int restore(struct ckpt_ctx *, struct ckpt_hdr_file *)
+ For most file systems, generic_file_{checkpoint,restore}() can be
+ used.
+
+* To checkpoint a socket, the 'struct proto_ops' needs to provide
+ the methods ->checkpoint, ->collect and ->restore:
+ int checkpoint(struct ckpt_ctx *ctx, struct socket *sock);
+ int collect(struct ckpt_ctx *ctx, struct socket *sock);
+ int restore(struct ckpt_ctx *, struct socket *sock, struct ckpt_hdr_socket *h)
+
diff --git a/Documentation/checkpoint/self_checkpoint.c b/Documentation/checkpoint/self_checkpoint.c
new file mode 100644
index 0000000..27dba0d
--- /dev/null
+++ b/Documentation/checkpoint/self_checkpoint.c
@@ -0,0 +1,69 @@
+/*
+ * self_checkpoint.c: demonstrate self-checkpoint
+ *
+ * Copyright (C) 2008 Oren Laadan
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file COPYING in the main directory of the Linux
+ * distribution for more details.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <errno.h>
+#include <math.h>
+#include <sys/syscall.h>
+
+#include <linux/checkpoint.h>
+
+static inline int checkpoint(pid_t pid, int fd, unsigned long flags)
+{
+ return syscall(__NR_checkpoint, pid, fd, flags, CHECKPOINT_FD_NONE);
+}
+
+#define OUTFILE "/tmp/cr-self.out"
+
+int main(int argc, char *argv[])
+{
+ pid_t pid = getpid();
+ FILE *file;
+ int i, ret;
+
+ close(0);
+ close(2);
+
+ unlink(OUTFILE);
+ file = fopen(OUTFILE, "w+");
+ if (!file) {
+ perror("open");
+ exit(1);
+ }
+ if (dup2(0, 2) < 0) {
+ perror("dup2");
+ exit(1);
+ }
+
+ fprintf(file, "hello, world!\n");
+ fflush(file);
+
+ for (i = 0; i < 1000; i++) {
+ sleep(1);
+ fprintf(file, "count %d\n", i);
+ fflush(file);
+
+ if (i != 2)
+ continue;
+ ret = checkpoint(pid, STDOUT_FILENO, CHECKPOINT_SUBTREE);
+ if (ret < 0) {
+ fprintf(file, "ckpt: %s\n", strerror(errno));
+ exit(2);
+ }
+
+ fprintf(file, "checkpoint ret: %d\n", ret);
+ fflush(file);
+ }
+
+ return 0;
+}
diff --git a/Documentation/checkpoint/self_restart.c b/Documentation/checkpoint/self_restart.c
new file mode 100644
index 0000000..647ce51
--- /dev/null
+++ b/Documentation/checkpoint/self_restart.c
@@ -0,0 +1,40 @@
+/*
+ * self_restart.c: demonstrate self-restart
+ *
+ * Copyright (C) 2008 Oren Laadan
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file COPYING in the main directory of the Linux
+ * distribution for more details.
+ */
+
+#define _GNU_SOURCE /* or _BSD_SOURCE or _SVID_SOURCE */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <unistd.h>
+#include <unistd.h>
+#include <sys/syscall.h>
+
+#include <linux/checkpoint.h>
+
+static inline int restart(pid_t pid, int fd, unsigned long flags)
+{
+ return syscall(__NR_restart, pid, fd, flags, CHECKPOINT_FD_NONE);
+}
+
+int main(int argc, char *argv[])
+{
+ pid_t pid = getpid();
+ int ret;
+
+ ret = restart(pid, STDIN_FILENO, RESTART_TASKSELF);
+ if (ret < 0)
+ perror("restart");
+
+ printf("should not reach here !\n");
+
+ return 0;
+}
diff --git a/Documentation/checkpoint/usage.txt b/Documentation/checkpoint/usage.txt
new file mode 100644
index 0000000..c6fc045
--- /dev/null
+++ b/Documentation/checkpoint/usage.txt
@@ -0,0 +1,247 @@
+
+ How to use Checkpoint-Restart
+ =========================================
+
+
+API
+===
+
+The API consists of three new system calls:
+
+* long checkpoint(pid_t pid, int fd, unsigned long flag, int logfd);
+
+ Checkpoint a (sub-)container whose root task is identified by @pid,
+ to the open file indicated by @fd. If @logfd isn't -1, it indicates
+ an open file to which error and debug messages are written. @flags
+ may be one or more of:
+ - CHECKPOINT_SUBTREE : allow checkpoint of sub-container
+ (other value are not allowed).
+
+ Returns: a positive checkpoint identifier (ckptid) upon success, 0 if
+ it returns from a restart, and -1 if an error occurs. The ckptid will
+ uniquely identify a checkpoint image, for as long as the checkpoint
+ is kept in the kernel (e.g. if one wishes to keep a checkpoint, or a
+ partial checkpoint, residing in kernel memory).
+
+* long sys_restart(pid_t pid, int fd, unsigned long flags, int logfd);
+
+ Restart a process hierarchy from a checkpoint image that is read from
+ the blob stored in the file indicated by @fd. If @logfd isn't -1, it
+ indicates an open file to which error and debug messages are written.
+ @flags will have future meaning (must be 0 for now). @pid indicates
+ the root of the hierarchy as seen in the coordinator's pid-namespace,
+ and is expected to be a child of the coordinator. @flags may be one
+ or more of:
+ - RESTART_TASKSELF : (self) restart of a single process
+ - RESTART_FROEZN : processes remain frozen once restart completes
+ - RESTART_GHOST : process is a ghost (placeholder for a pid)
+ (Note that this argument may mean 'ckptid' to identify an in-kernel
+ checkpoint image, with some @flags in the future).
+
+ Returns: -1 if an error occurs, 0 on success when restarting from a
+ "self" checkpoint, and return value of system call at the time of the
+ checkpoint when restarting from an "external" checkpoint.
+
+ (If a process was frozen for checkpoint while in userspace, it will
+ resume running in userspace exactly where it was interrupted. If it
+ was frozen while in kernel doing a syscall, it will return what the
+ syscall returned when interrupted/completed, and proceed from there
+ as if it had only been frozen and then thawed. Finally, if it did a
+ self-checkpoint, it will resume to the first instruction after the
+ call to checkpoint(2), having returned 0, to indicate whether the
+ return is from the checkpoint or a restart).
+
+* int clone_with_pid(unsigned long clone_flags, void *news,
+ int *parent_tidptr, int *child_tidptr,
+ struct target_pid_set *pid_set)
+
+ struct target_pid_set {
+ int num_pids;
+ pid_t *target_pids;
+ }
+
+ Container restart requires that a task have the same pid it had when
+ it was checkpointed. When containers are nested the tasks within the
+ containers exist in multiple pid namespaces and hence have multiple
+ pids to specify during restart.
+
+ clone_with_pids(), intended for use during restart, is similar to
+ clone(), except that it takes a 'target_pid_set' parameter. This
+ parameter lets caller choose specific pid numbers for the child
+ process, in the process's active and ancestor pid namespaces.
+
+ Unlike clone(), clone_with_pids() needs CAP_SYS_ADMIN, at least for
+ now, to prevent unprivileged processes from misusing this interface.
+
+ If a target-pid is 0, the kernel continues to assign a pid for the
+ process in that namespace. If a requested pid is taken, the system
+ call fails with -EBUSY. If 'pid_set.num_pids' exceeds the current
+ nesting level of pid namespaces, the system call fails with -EINVAL.
+
+
+Sysctl/proc
+===========
+
+/proc/sys/kernel/ckpt_unpriv_allowed [default = 1]
+ controls whether c/r operation is allowed for unprivileged users
+
+
+Operation
+=========
+
+The granularity of a checkpoint usually is a process hierarchy. The
+'pid' argument is interpreted in the caller's pid namespace. So to
+checkpoint a container whose init task (pid 1 in that pidns) appears
+as pid 3497 the caller's pidns, the caller must use pid 3497. Passing
+pid 1 will attempt to checkpoint the caller's container, and if the
+caller isn't privileged and init is owned by root, it will fail.
+
+Unless the CHECKPOINT_SUBTREE flag is set, if the caller passes a pid
+which does not refer to a container's init task, then sys_checkpoint()
+would return -EINVAL.
+
+We assume that during checkpoint and restart the container state is
+quiescent. During checkpoint, this means that all affected tasks are
+frozen (or otherwise stopped). During restart, this means that all
+affected tasks are executing the sys_restart() call. In both cases, if
+there are other tasks possible sharing state with the container, they
+must not modify it during the operation. It is the responsibility of
+the caller to follow this requirement.
+
+If the assumption that all tasks are frozen and that there is no other
+sharing doesn't hold - then the results of the operation are undefined
+(just as, e.g. not calling execve() immediately after vfork() produces
+undefined results). In particular, either checkpoint will fail, or it
+may produce a checkpoint image that can't be restarted, or (unlikely)
+the restart may produce a container whose state does not match that of
+the original container.
+
+
+User tools
+==========
+
+* checkpoint(1): a tool to perform a checkpoint of a container/subtree
+* restart(1): a tool to restart a container/subtree
+* ckptinfo: a tool to examine a checkpoint image
+
+It is best to use the dedicated user tools for checkpoint and restart.
+
+If you insist, then here is a code snippet that illustrates how a
+checkpoint is initiated by a process inside a container - the logic is
+similar to fork():
+ ...
+ ckptid = checkpoint(0, ...);
+ switch (crid) {
+ case -1:
+ perror("checkpoint failed");
+ break;
+ default:
+ fprintf(stderr, "checkpoint succeeded, CRID=%d\n", ret);
+ /* proceed with execution after checkpoint */
+ ...
+ break;
+ case 0:
+ fprintf(stderr, "returned after restart\n");
+ /* proceed with action required following a restart */
+ ...
+ break;
+ }
+ ...
+
+And to initiate a restart, the process in an empty container can use
+logic similar to execve():
+ ...
+ if (restart(pid, ...) < 0)
+ perror("restart failed");
+ /* only get here if restart failed */
+ ...
+
+Note, that the code also supports "self" checkpoint, where a process
+can checkpoint itself. This mode does not capture the relationships of
+the task with other tasks, or any shared resources. It is useful for
+application that wish to be able to save and restore their state.
+They will either not use (or care about) shared resources, or they
+will be aware of the operations and adapt suitably after a restart.
+The code above can also be used for "self" checkpoint.
+
+
+You may find the following sample programs useful:
+
+* checkpoint.c: accepts a 'pid' and checkpoint that task to stdout
+* self_checkpoint.c: a simple test program doing self-checkpoint
+* self_restart.c: restarts a (self-) checkpoint image from stdin
+
+See also the utilities 'checkpoint' and 'restart' (from user-cr).
+
+
+"External" checkpoint
+=====================
+
+To do "external" checkpoint, you need to first freeze that other task
+either using the freezer cgroup.
+
+Restart does not preserve the original PID yet, (because we haven't
+solved yet the fork-with-specific-pid issue). In a real scenario, you
+probably want to first create a new names space, and have the init
+task there call 'sys_restart()'.
+
+I tested it this way:
+ $ ./test &
+ [1] 3493
+
+ $ echo 3493 > /cgroup/0/tasks
+ $ echo FROZEN > /cgroup/0/freezer.state
+ $ ./checkpoint 3493 > ckpt.image
+
+ $ mv /tmp/cr-test.out /tmp/cr-test.out.orig
+ $ cp /tmp/cr-test.out.orig /tmp/cr-test.out
+
+ $ echo THAWED > /cgroup/0/freezer.state
+
+ $ ./self_restart < ckpt.image
+Now compare the output of the two output files.
+
+
+"Self" checkpoint
+================
+
+To do self-checkpoint, you can incorporate the code from
+self_checkpoint.c into your application.
+
+Here is how to test the self-checkpoint:
+ $ ./self_checkpoint > self.image &
+ [1] 3512
+
+ $ sleep 3
+ $ mv /tmp/cr-self.out /tmp/cr-self.out.orig
+ $ cp /tmp/cr-self.out.orig /tmp/cr-self.out
+
+ $ cat /tmp/cr-self.out
+ hello, world!
+ count 0
+ count 1
+ count 2
+ checkpoint ret: 1
+ count 3
+ ...
+
+ $ sed -i 's/count/xxxxx/g' /tmp/cr-self.out
+
+ $ ./self_restart < self.image &
+
+Now compare the output of the two output files.
+ $ cat /tmp/cr-self.out
+ hello, world!
+ xxxxx 0
+ xxxxx 1
+ xxxxx 2
+ checkpoint ret: 0
+ count 3
+ ...
+
+
+Note how in test.c we close stdin, stdout, stderr - that's because
+currently we only support regular files (not ttys/ptys).
+
+If you check the output of ps, you'll see that "self_restart" changed
+its name to "test" or "self_checkpoint", as expected.
--
1.6.3.3
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>