Ralph Castain 86ab751c5e Next step in reducing launch time: begin reducing the size of the launch message itself. Start by expressing the daemon map as a set of three regular expression strings. On an 8k cluster, this reduces the nidmap contribution from over 200kBytes to 21 bytes in size. ...

Signed-off-by: Ralph Castain <rhc@open-mpi.org>

2017-01-23 19:54:47 -08:00

..

alps

pmix/cray: fix disable-dlopen problem

2016-11-21 13:45:10 -06:00

base

Next step in reducing launch time: begin reducing the size of the launch message itself. Start by expressing the daemon map as a set of three regular expression strings. On an 8k cluster, this reduces the nidmap contribution from over 200kBytes to 21 bytes in size.

2017-01-23 19:54:47 -08:00

default

Update signal handling to introduce a pause between SIGCONT and SIGTERM, followed by another pause before SIGKILL. Do this within the odls/kill_local_procs function while we know we are blocked in an event, and before the daemon shuts down the event progress loop

2016-12-06 12:34:42 -08:00

Makefile.am

Purge whitespace from the repo

2015-06-23 20:59:57 -07:00

odls_types.h

Revert e9bc293 and try a different approach for scalably dealing with hetero clusters. Have each orted send back its topo "signature". If mpirun detects that this signature has not been seen before, then ask for that daemon to send back its full topology description. This allows the system to only get the topology once for each unique topo in the cluster.

2017-01-18 10:22:15 -08:00

odls.h

Refactor the ORTE DVM code so that external codes can submit multiple jobs using only a single connection to the HNP.

2016-02-13 08:10:44 -08:00