DMTCP(1) Distributed MultiThreaded CheckPointing DMTCP(1) (NAME) dmtcp - (Distributed MultiThreaded Checkpointing) (SYNOPSIS) dmtcp_coordinator [port] dmtcp_launch command [args...] dmtcp_restart ckpt_FILE1.dmtcp [ckpt_FILE2.dmtcp...] dmtcp_command coordinatorCommand (DESCRIPTION) DMTCP ( ) . . MTCP DMTCP . (OPTIONS) --help -h . . bash "export NAME=value" tcsh "setenv NAME value" . DMTCP_CHECKPOINT_INTERVAL=integer . 'c' . (: 0 dmtcp_coordinator) DMTCP_HOST=string . (: localhost dmtcp_launch dmtcp_restart) DMTCP_PORT=integer . (: 7779) DMTCP_GZIP=(1|0) "0" . (: 1 dmtcp_launch) : gzip . gzip (ckpt/restart) . DMTCP_CHECKPOINT_DIR=path . (: ./) DMTCP_SIGCKPT=integer . . (: SIGUSR2 dmtcp_launch) DMTCP (DMTCP_COORDINATOR) DMTCP . dmtcp_coordinator dmtcp_launch dmtcp_restart . dmtcp_launch dmtcp_restart dmtcp_command --host --port DMTCP_HOST DMTCP_PORT . - localhost-7779 . - dmtcp_coordinator localhost-7779 . (stateless) . (restart) . DMTCP ( - ) . . : ( DMTCP_CHECKPOINT_INTERVAL ) ( ) (plugin API) ( ). . . : l : s : c : ( ) f : () k : (kill) q : ? : dmtcp_command . (EXAMPLE USAGE) 1. dmtcp_coordinator . ( .) dmtcp_coordinator 2. "dmtcp_launch [command]" . DMTCP_HOST DMTCP_PORT . . . ssh . ( DMTCP ssh dmtcp_launch .) dmtcp_launch ./myprogram 3. "c" . DMTCP_CHECKPOINT_DIR . dmtcp_coordinator "dmtcp_restart_script.sh" . dmtcp_restart ssh . dmtcp_command -c : dmtcp_command --checkpoint 4. dmtcp_restart_script.sh dmtcp_coordinator . . . . ./dmtcp_restart_script.sh DMTCP (DMTCP PLUGIN) plugin DMTCP . . test/plugin . test/plugin/applic-initiated . . : (wrappers) ( / ) - . ( dmtcpaware .) : dmtcp_launch --with-plugin PLUGIN1[:PLUGIN2]... DMTCPAWARE (DMTCPAWARE API) DMTCP ( ). DMTCP dmtcp . dmtcpaware/dmtcpaware.h . test/dmtcpaware[123].c . : cd test; rm dmtcpaware1; make dmtcpaware1; ./autotest -v dmtcpaware1 libdmtcpaware.so (-ldmtcpaware) dmtcp/dmtcpaware.h . (RETURN CODE) DMTCP DMTCP . DMTCP ( ) DMTCP rc ( rc+1 rc+2) rc DMTCP_FAIL_RC 99 . (SEE ALSO) DMTCP : http://dmtcp.sourceforge.net (FAQ): http://dmtcp.sourceforge.net/FAQ.html : https://github.com/dmtcp/dmtcp/blob/master/doc/plugin-tutorial.pdf (DMTCP-2.x) : https://github.com/dmtcp/dmtcp/tree/master/test/plugin (AUTHORS) DMTCP MTCP (MultiThreaded CheckPointing) Jason Ansel Kapil Arya Gene Cooperman Rohan Garg Artem Y. Polyakov Mike Rieker Alex Brick Tyler Denniston William Enright Gregory Kerr Ana-Maria Visan . DMTCP . DMTCP team June 17, 2008 DMTCP(1)