/usr/src/blcr-0.8.5/examples/README is in blcr-dkms 0.8.5-2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 | This directory contains a number of example programs, each in its
own directory with its own Makefile.
NOTE: These are not "supported" in the sense of providing any particular
value to you. We provide these in case they might be usefull. These
examples are not written as templates for developing BLCR client codes.
However, pthread_misc and io_bench do contain some of the basics.
We hope that in the future we can revise and expand the examples to
make them suitable as templates for developing BLCR client codes.
1) Example checkpoint "targets".
These are a few simple variations on a theme that can be used to
play with BLCR. You run an example and can checkpoint and restart
it seeing that BLCR works as expected.
* counting
This is a simple program that counts from 0 to 119, with 1 second
pauses between each count. When restarting it will go back to the
count value when the checkpoint was taken.
* pipe_counting
This program is a more complicated version of the basic counting
example. The process fork()s a child which does the printing of
counts. The actual count values are passed from parent to child
through a pipe. So, this example demonstates that BLCR will
restore that pipe when checkpointing the two processes together.
* pthread_counting
This program is a multi-threaded (using pthreads) variation on
the basic counting example. In this case there are 3 threads that
take turns producing output. So, this demonstrates that BLCR is
able to restore multi-threaded apps and their pthread synchronization
primitives (a mutex and a condition variable in this case).
* file_counting
This example is similar to "counting", except that the output is
to the file "outfile". Because BLCR records the length of the file
when checkpointing, a restart will rewind the file back to its
pre-checkpoint length at each restart. Therefore the file will
never contain duplicate counts.
2) Example BLCR "clients".
* pthread_misc
This example initializes the BLCR library and registers a callback
to be invoked at checkpoint time. In this simple example the
callback is simply used to record a timestamp for the checkpoint,
which is printed at restart.
3) Example benchmark
* io_bench
This program takes a single argument: a size in MiB. It allocates
(and touches) the corresponding amount of memory and then takes a
checkpoint of itself. It reports the allocated heap size and the
amount of time required for the checkpoint.
|