HP-MPI V2.3 for Linux Release Note

another HP-MPI job while the first job remains suspended. When a suspended mpirun
job receives a SIGCONT, the licenses are reacquired and the job continues. If the licenses
cannot be reacquired from the license server, the job exits.
NOTE: When a job is suspended in Linux, any memory that is pinned is not swapped
to disk, and is not handled by the operating system's virtual memory subsystem.
HP-MPI pins memory that is associated with RDMA message transfers. By default, up
to 20% of the system memory can be pinned by HP-MPI at any one time. The amount
of memory that is pinned can be changed by two environment variables:
MPI_PHYSICAL_MEMORY and MPI_PIN_PERCENTAGE (default 20%). The -dd option
to mpirun displays the amount of physical memory that is detected by HP-MPI. If the
detection is wrong, the correct amount of physical memory should be set with
MPI_PHYSICAL_MEMORY in bytes. This memory is only returned to the OS for use by
other processes after the job resumes and exits.
1.2.7.7 Expanded Functionality for -ha
HP-MPI high availability mode eliminates teardown when ranks exit abnormally.
Further communications involved with ranks that are unreachable return error class
MPI_ERR_EXITED, but do not force the application to teardown as long as the
MPI_Errhandler is set to MPI_ERRORS_RETURN. This mode never uses shared
memory for inter-process communication.
HP-MPI high availability mode is accessed by using the -ha option to mpirun. The
high availability mode has been expanded substantially in this release. Some of the
expanded features are available only on HP platforms.
To allow users to select the correct level of high availability features for an application,
the -ha option now accepts a number of additional colon-separated options which
may be appended to the -ha command line option. For example:
mpirun -ha[:option1][:option2][...]
Table 1-5 High Availability Options
DescriptionsOptions
Basic high availability protection. When specified with no options, -ha
is equivalent to -ha:noteardown:detect.
-ha
Use of lightweight instrumentation with -ha. HP hardware only.-ha -i
High availability for infrastructure (mpirun, mpid). HP hardware only.
For more information, see “Highly Available Infrastructure (-ha:infra)”
(page 16).
-ha:infra
Detection of failed communication connections. For more information,
see “Failure Detection (-ha:detect)” (page 19).
-ha:detect
1.2 What’s in This Version 15