Home | Downloads | Issues | Bug reporting

SGE Checkpointing with DMTCP

There are (at least) two versions of support for checkpointing with DMTCP under SGE:
for use as a starter method;
dmtcpckpt script
potentially usable with qsub -S or extendable as a starter method; otherwise to be execed in a job taking account of SGE's RESTARTED variable and maybe setting a restart exit code. NB. not properly tested in production. See its --help option or the man page.