|
mca
|
mca_param_files
|
Path for MCA configuration files containing default parameter values
|
$HOME/.openmpi/mca-params.conf:/opt/SUNWhpc/HPC8.1/etc/openmpi-mca-params.conf
|
|
|
mca_component_path
|
Path where to look for Open MPI and ORTE components
|
/opt/SUNWhpc/HPC8.1/lib/openmpi:$HOME/.openmpi/components
|
|
|
mca_verbose
|
Top-level verbosity parameter
|
none
|
|
|
mca_component_show_load_errors
|
Whether or not to show errors for components that failed to load
|
1
|
|
|
mca_component_disable_dlopen
|
Whether to attempt to disable opening dynamic components or not
|
0
|
|
mpi
|
mpi_param_check
|
Whether you want MPI API parameters checked at run-time or not. Possible values are 0 (no checking) and 1 (perform checking at runtime)
|
1
|
|
|
mpi_yield_when_idle
|
Yield the processor when waiting for MPI communication (for MPI processes, will default to 1 when oversubscribing nodes)
|
0
|
|
|
mpi_event_tick_rate
|
How often to progress TCP communications (-1 = leave default value. 0 = never, otherwise specified in microseconds)
|
10000
|
|
|
mpi_show_handle_leaks
|
Whether or not MPI_FINALIZE shows all MPI handles that were
not freed
|
0
|
|
|
mpi_no_free_handles
|
Whether to actually free MPI objects when their handles are freed
|
0
|
|
|
mpi_show_mca_params
|
Whether to show all MCA parameter value during MPI_INIT or not (good for reproducibility of MPI jobs)
|
0 (don’t show)
|
|
|
mpi_show_mca_params_file
|
If mpi_show_mca_params is true, setting this string to a valid file name tells Open MPI to dump all the MCA parameter values into a file suitable for reading using the mca_param_files parameter (good for reproducibility of MPI jobs
|
none
|
|
|
opal_paffinity_alone
|
If nonzero, assume that this job is the only (set of) process(es) running on each node and bind processes to processors, starting with processor ID 0
|
0
|
|
|
mpi_keep_peer_hostnames
|
If nonzero, save the string hostnames of all MPI peer processes (mostly for error / debugging output messages). This can add quite a bit of memory usage to each MPI process.
|
1
|
|
|
mpi_abort_delay
|
If nonzero, print out an identifying message when MPI_ABORT is invoked (hostname, PID of the process that called MPI_ABORT) and delay for that many seconds before exiting (a negative delay value means to never abort). This allows attaching of a debugger before quitting the job.
|
0
|
|
|
mpi_abort_print_stack
|
If nonzero, print out a stack trace when MPI_ABORT is invoked
|
0
Note: this is an information parameter. Its value is returned by Open MPI. This value cannot be adjusted manually.
|
|
|
mpi_preconnect_all
|
Whether to force MPI processes to create connections/ warmup with all peers during MPI_INIT (vs. making connections lazily -- upon the first MPI traffic between each process/peer pair
|
0
|
|
|
mpi_leave_pinned
|
Whether to use the "leave pinned" protocol or not. Enabling this setting can help bandwidth performance when repeatedly sending and receiving large messages with the same buffers over RDMA-based networks. Use this parameter with care; setting it to too high a value can lead to resource exhaustion.
|
0
|
|
|
mpi_leave_pinned_pipeline
|
Whether to use the "leave pinned pipeline" protocol or not.
|
0
|
|
orte
|
orte_base_user_debugger
|
Sequence of user-level debuggers to search for in orterun
|
totalview @mpirun@ -a @mpirun_args@ : fxp @mpirun@ -a @mpirun_args@
|
|
|
orte_debug
|
Whether or not to enable debugging output for all ORTE components (0 or 1)
|
0
|
|
|
orte_no_daemonize
|
Whether to properly daemonize the ORTE daemons or not
|
0
|
|
|
orte_abort_timeout
|
Time to wait [in seconds] before giving up on aborting an ORTE operation
|
10
|
|
|
orte_timing
|
Request that critical timing loops be measured
|
0
|
|
opal
|
opal_signal
|
If a signal is received, display the stack trace frame
|
6,10,8,11
|
|
backtrace
|
backtrace
|
Default selection set of components for the backtrace framework (<none> means "use all components that can be found"
|
none
|
|
|
backtrace_base_verbose
|
Verbosity level for the backtrace framework (0 = no verbosity)
|
0
|
|
|
backtrace_printstack_priority
|
|
0
|
|
memory
|
memory
|
Default selection set of components for the memory framework (none means "use all components that can be found"
|
none
|
|
|
memory_base_verbose
|
Verbosity level for the memory framework (0 = no verbosity)
|
0
|
|
paffinity
|
paffinity
|
Default selection set of components for the paffinity framework (none means "use all components that can be found”)
|
none
|
|
|
paffinity_solaris_priority
|
Priority of the Solaris OS paffinity component
|
10
|
|
maffinity
|
maffinity
|
Default selection set of components for the maffinity framework (none means "use all components that can be found")
|
none
|
|
|
maffinity_first_use_priority
|
Priority of the first_use maffinity component
|
10
|
|
timer
|
timer
|
Default selection set of components for the timer framework (<none> means "use all components that can be found")
|
none
|
|
|
timer_base_verbose
|
Verbosity level for the timer framework (0 = no verbosity)
|
0
|
|
|
timer_solaris_priority
|
Priority of the Solaris OS maffinity component
|
0
|
|
allocator
|
allocator
|
Default selection set of components for the allocator framework (none means "use all components that can be found")
|
none
|
|
|
allocator_base_verbose
|
Verbosity level for the allocator framework (0 = no verbosity)
|
0
|
|
|
allocator_basic_priority
|
Priority of the basic allocator component
|
0
|
|
|
allocator_bucket_num_buckets
|
Number of buckets in the allocator bucket component
|
30
|
|
|
allocator_bucket_priority
|
Priority of the allocator bucket component
|
0
|
|
coll
|
coll
|
Default selection set of components for the coll framework (none means "use all components that can be found")
|
none
|
|
|
coll_base_verbose
|
Verbosity level for the coll framework (0 = no verbosity)
|
0
|
|
|
coll_basic_priority
|
Priority of the basic coll component
|
10
|
|
|
coll_basic_crossover
|
Minimum number of processes in a communicator before
using the logarithmic algorithms
|
4
|
|
|
coll_self_priority
|
Priority of the self coll component
|
75
|
|
|
coll_sm_priority
|
Priority of the sm coll component
|
0
|
|
|
coll_sm_control_size
|
Length of the control data -- should usually be either the length of a cache line on most multiprocessors, or the size of a page on machines that support direct memory affinity page placement (in bytes)
|
4096
|
|
|
coll_sm_bootstrap_filename
|
File name (in the Open MPI session directory) of the coll sm component bootstrap rendezvous mmap file
|
shared_mem_sm_
bootstrap
|
|
|
coll_sm_bootstrap_num_segments
|
Number of segments in the bootstrap file
|
8
|
|
|
coll_sm_fragment_size
|
Fragment size (in bytes) used for passing data through shared memory (will be rounded up to the nearest control_size size)
|
8192
|
|
|
coll_sm_mpool
|
Name of the mpool component to use
|
sm
|
|
|
coll_sm_comm_in_use_flags
|
Number of "in use" flags, used to mark a message passing area segment as currently being used or not (must be >= 2 and <= comm_num_segments)
|
2
|
|
|
coll_sm_comm_num_segments
|
Number of segments in each communicator’s shared memory message passing area (must be >= 2, and must be a multiple of comm_in_use_flags)
|
8
|
|
|
coll_sm_tree_degree
|
Degree of the tree for tree-based operations (must be => 1 and <= min(control_size, 255))
|
4
|
|
|
coll_sm_shared_mem_used_
bootstrap
|
Amount of shared memory used in the shared memory bootstrap area (in bytes)
|
160
Note: this is an information parameter. Its value is returned by Open MPI. This value cannot be adjusted manually.
|
|
|
coll_sm_info_num_procs
|
Number of processes to use for the calculation of the shared_mem_size MCA information parameter (must be => 2)
|
4
|
|
|
coll_sm_shared_mem_used_data
|
Amount of shared memory used in the shared memory data area for info_num_procs processes (in bytes)
|
548864
Note: this is an information parameter. Its value is returned by Open MPI. This value cannot be adjusted manually.
|
|
|
coll_tuned_priority
|
Priority of the tuned coll component
|
30
|
|
|
coll_tuned_pre_allocate_memory_comm_size_limit
|
Size of communicator where you want to stop pre-allocating memory for the fixed internal buffer. The buffer is used for message requests attached to the communicator data segment. For example, if you have a 100,000-node cluster, you might not
want to pre-allocate 200,000 request handle slots per communicator instance.
|
32768
|
|
|
coll_tuned_use_dynamic_rules
|
Switch used to decide whether to use static (compiled/if statements) or dynamic (built at runtime) decision function rules
|
0
|
|
|
coll_tuned_init_tree_fanout
|
Initial fanout used in the tree topologies for each communicator. This is only an initial estimate value. If a tuned collective needs a different fanout for an operation, it builds it dynamically. This parameter is only for the first guess and might save a little time.
|
4
|
|
|
coll_tuned_init_chain_fanout
|
Initial fanout used in the chain (fanout followed by pipeline) topologies for each communicator. This is only an initial estimate. If a tuned collective needs a different fanout for an operation, it builds it dynamically. This parameter is only for the first guess and might save a little time.
|
4
|
|
|
coll_tuned_use_dynamic_rules
|
Switch used to decide whether to use static (compiled/if statements) or dynamic (built at runtime) decision function rules (0 = use static rules)
|
0
|
|
io
|
io
|
Default selection set of components for the io framework (none means "use all components that can be found")
|
none
|
|
|
io_base_freelist_initial_size
|
Initial MPI-2 IO request freelist size
|
16
|
|
|
io_base_freelist_max_size
|
Max size of the MPI-2 IO request freelist
|
64
|
|
|
io_base_freelist_increment
|
Increment size of the MPI-2 IO request freelist
|
16
|
|
|
io_base_verbose
|
Verbosity level for the io framework (0 = no verbosity)
|
0
|
|
|
io_romio_priority
|
Priority of the io romio component
|
10
|
|
|
io_romio_delete_priority
|
Delete priority of the io romio component
|
10
|
|
|
io_romio_enable_parallel_
optimizations
|
Enable set of Open MPI-added options to improve collective file I/O performance
|
0
|
|
mpool
|
mpool
|
Default selection set of components for the mpool framework (<none> means "use all components that can be found")
|
none
|
|
|
mpool_base_verbose
|
Verbosity level for the mpool framework (0 = no verbosity)
|
0
|
|
|
mpool_sm_allocator
|
Type of shared memory allocator used
|
bucket
|
|
|
mpool_sm_max_size
|
Maximum size of the sm mpool shared memory file
|
536870912
|
|
|
mpool_sm_min_size
|
Minimum size of the sm mpool shared memory file
|
134217728
|
|
|
mpool_sm_per_peer_size
|
Size (in bytes) to allocate per local peer in the sm mpool shared memory file, bounded by min_size and max_size
|
33554432
|
|
|
mpool_sm_priority
|
Priority of the shared memory allocator for mpool
|
0
|
|
|
mpool_udapl_priority
|
Priority of the uDAPL BTL
|
0
|
|
|
mpool_base_use_mem_hooks
|
Use memory hooks for deregistering freed memory
|
0
|
|
|
mpool_use_mem_hooks
|
Deprecated; use mpool_base_use_mem_
hook
|
0
|
|
pml
|
pml
|
Default selection set of components for the pml framework (none means "use all components that can be found")
|
ob1
|
|
|
pml_base_verbose
|
Verbosity level for the pml framework (0 = no verbosity)
|
0
|
|
|
pml_cm_free_list_num
|
Initial size of request free lists
|
4
|
|
|
pml_cm_free_list_max
|
Maximum size of request free lists
|
-1
|
|
|
pml_cm_free_list_inc
|
Number of elements to add when growing request free lists
|
64
|
|
|
pml_cm_priority
|
CM PML selection priority
|
1
|
|
|
pml_dr_free_list_num
|
|
4
|
|
|
pml_dr_free_list_max
|
|
-1
|
|
|
pml_dr_free_list_inc
|
|
64
|
|
|
pml_dr_priority
|
|
1
|
|
|
pml_dr_eager_limit
|
|
131072
|
|
|
pml_dr_send_pipeline_depth
|
|
3
|
|
|
pml_dr_wdog_timer_sec
|
|
5
|
|
|
pml_dr_wdog_timer_usec
|
|
0
|
|
|
pml_dr_wdog_timer_multiplier
|
|
1
|
|
|
pml_dr_wdog_retry_max
|
|
1
|
|
|
pml_dr_ack_timer_sec
|
|
10
|
|
|
pml_dr_ack_timer_usec
|
|
0
|
|
|
pml_dr_ack_timer_multiplier
|
|
1
|
|
|
pml_dr_ack_retry_max
|
|
3
|
|
|
pml_dr_enable_csum
|
|
1
|
|
|
pml_ob1_free_list_num
|
|
4
|
|
|
pml_ob1_free_list_max
|
|
-1
|
|
|
pml_ob1_free_list_inc
|
|
64
|
|
|
pml_ob1_priority
|
|
1
|
|
|
pml_ob1_eager_limit
|
|
131072
|
|
|
pml_ob1_send_pipeline_depth
|
|
3
|
|
|
pml_ob1_recv_pipeline_depth
|
|
4
|
|
bml
|
bml
|
Default selection set of components for the bml framework (none means "use all components that can be found")
|
none
|
|
|
bml_base_verbose
|
Verbosity level for the bml framework (0 = no verbosity)
|
0
|
|
|
bml_r2_show_unreach_errors
|
Show error message when procs are unreachable
|
1
|
|
|
bml_r2_priority
|
|
0
|
|
rcache
|
rcache
|
Default selection set of components for the rcache framework (none means "use all components that can be found")
|
none
|
|
|
rcache_base_verbose
|
Verbosity level for the rcache framework (0 = no verbosity)
|
0
|
|
|
rcache_rb_priority
|
|
0
|
|
|
rcache_vma_mru_len
|
The maximum size IN ENTRIES of the MRU (most recently used) rcache list
|
256
|
|
|
rcache_vma_mru_size
|
The maximum size IN BYTES of the MRU (most recently used) rcache list
|
1073741824
|
|
|
rcache_vma_priority
|
|
0
|
|
btl
|
btl
|
Default selection set of components for the btl framework (none means "use all components that can be found")
|
none
|
|
|
btl_base_debug
|
If btl_base_debug is 1 standard debug is output, if > 1 verbose debug is output
|
0
|
|
|
btl_base_verbose
|
Verbosity level for the btl framework (0 = no verbosity)
|
0
|
|
|
btl_self_free_list_num
|
Number of fragments by default
|
0
|
|
|
btl_self_free_list_max
|
Maximum number of fragments
|
-1
|
|
|
btl_self_free_list_inc
|
Increment by this number of fragments
|
32
|
|
|
btl_self_eager_limit
|
Eager size fragment (before the rendezvous protocol)
|
131072
|
|
|
btl_self_min_send_size
|
Minimum fragment size after the rendezvous
|
262144
|
|
|
btl_self_max_send_size
|
Maximum fragment size after the rendezvous
|
262144
|
|
|
btl_self_min_rdma_size
|
Minimum fragment size for the RDMA transfer
|
2147483647
|
|
|
btl_self_max_rdma_size
|
Maximum fragment size for the RDMA transfer
|
2147483647
|
|
|
btl_self_exclusivity
|
Device exclusivity
|
65536
|
|
|
btl_self_flags
|
Active behavior flags
|
10
|
|
|
btl_self_priority
|
Device priority
|
0
|
|
|
btl_sm_free_list_num
|
|
8
|
|
|
btl_sm_free_list_max
|
|
-1
|
|
|
btl_sm_free_list_inc
|
|
256
|
|
|
btl_sm_exclusivity
|
|
65535
|
|
|
btl_sm_latency
|
|
100
|
|
|
btl_sm_max_procs
|
|
-1
|
|
|
btl_sm_sm_extra_procs
|
|
2
|
|
|
btl_sm_mpool
|
|
sm
|
|
|
btl_sm_eager_limit
|
|
4096
|
|
|
btl_sm_max_frag_size
|
|
32768
|
|
|
btl_sm_size_of_cb_queue
|
|
128
|
|
|
btl_sm_cb_lazy_free_freq
|
|
120
|
|
|
btl_sm_priority
|
|
0
|
|
|
btl_tcp_if_include
|
|
none
|
|
|
btl_tcp_if_exclude
|
|
lo
|
|
|
btl_tcp_free_list_num
|
|
8
|
|
|
btl_tcp_free_list_max
|
|
-1
|
|
|
btl_tcp_free_list_inc
|
|
32
|
|
|
btl_tcp_sndbuf
|
|
131072
|
|
|
btl_tcp_rcvbuf
|
|
131072
|
|
|
btl_tcp_endpoint_cache
|
|
30720
|
|
|
btl_tcp_exclusivity
|
|
0
|
|
|
btl_tcp_eager_limit
|
|
65536
|
|
|
btl_tcp_min_send_size
|
|
65536
|
|
|
btl_tcp_max_send_size
|
|
131072
|
|
|
btl_tcp_min_rdma_size
|
The user message size at which point the MPI library will begin registering portions of the user’s send and receive buffers, and begin using RDMA operations to write directly from user buffer to user buffer. The write operations occur in portions no greater than the value of btl_tcp_max_rdma_
size.
|
131072
|
|
|
btl_tcp_max_rdma_size
|
The maximum message size that the MPI library will use to perform the actual RDMA operation.
|
2147483647
|
|
|
btl_tcp_flags
|
|
10
|
|
|
btl_tcp_priority
|
|
0
|
|
|
btl_udapl_free_list_num
|
|
8
|
|
|
btl_udapl_free_list_max
|
|
-1
|
|
|
btl_udapl_free_list_inc
|
|
8
|
|
|
btl_udapl_mpool
|
|
udapl
|
|
|
btl_udapl_max_modules
|
|
8
|
|
|
btl_udapl_evd_qlen
|
|
32
|
|
|
btl_udapl_max_request_dtos
|
|
18
|
|
|
btl_udapl_max_recv_dtos
|
|
18
|
|
|
btl_udapl_num_recvs
|
|
8
|
|
|
btl_udapl_num_sends
|
|
8
|
|
|
btl_udapl_sr_win
|
|
4
|
|
|
btl_udapl_timeout
|
|
10000000
|
|
|
btl_udapl_eager_rdma_num
|
|
8
|
|
|
btl_udapl_max_eager_rdma_peers
|
|
16
|
|
|
btl_udapl_eager_rdma_win
|
|
4
|
|
|
btl_udapl_timeout
|
|
10000000
|
|
|
btl_udapl_eager_rdma_guarantee
|
|
0
|
|
|
btl_udapl_conn_priv_data
|
|
1
|
|
|
btl_udapl_exclusivity
|
|
1014
|
|
|
btl_udapl_eager_limit
|
|
32768
|
|
|
btl_udapl_min_send_size
|
|
16384
|
|
|
btl_udapl_max_send_size
|
|
65536
|
|
|
btl_udapl_min_rdma_size
|
The user message size at which point the MPI library will begin registering portions of the user’s send and receive buffers, and begin using RDMA operations to write directly from user buffer to user buffer. The write operations occur in portions no greater than the value of btl_udapl_max_rdma_
size.
|
524288
|
|
|
btl_udapl_max_rdma_size
|
The maximum message size that the MPI library will use to perform the actual RDMA operation.
|
131072
|
|
|
btl_udapl_flags
|
|
2
|
|
|
btl_udapl_bandwidth
|
|
225
|
|
|
btl_udapl_priority
|
|
0
|
|
|
btl_base_include
|
|
none
|
|
|
btl_base_exclude
|
|
none
|
|
|
btl_base_warn_component_unused
|
This parameter is used to turn on warning messages when certain NICs are not used.
|
0
|
|
mtl
|
mtl
|
Default selection set of components for the mtl framework (none means "use all components that can be found").
|
none
|
|
|
mtl_base_verbose
|
Verbosity level for the mtl framework (0 = no verbosity)
|
0
|
|
topo
|
topo
|
Default selection set of components for the topo framework (none means "use all components that can be found").
|
none
|
|
|
topo_base_verbose
|
Verbosity level for the topo framework (0 = no verbosity).
|
0
|
|
osc
|
osc
|
Default selection set of components for the osc framework (none means "use all components that can be found").
|
none
|
|
|
osc_base_verbose
|
Verbosity level for the osc framework (0 = no verbosity)
|
0
|
|
|
osc_pt2pt_no_locks
|
Enable optimizations available only if MPI_LOCK is not used.
|
0
|
|
|
osc_pt2pt_eager_limit
|
Max size of eagerly sent data
|
16384
|
|
|
osc_pt2pt_priority
|
|
0
|
|
|
osc_rdma_fence_sync_method
|
Specifies how to synchronize fence: reduce_scatter, allreduce,
alltoall
|
reduce_scatter
|
|
|
osc_rdma_eager_send
|
Attempt to start data movement during communication call, instead of at synchronization time. Info key of same name overrides this value, if info key given.
|
0
|
|
|
osc_rdma_no_locks
|
Enable optimizations available only if MPI_LOCK is not used.
|
0
|
|
|
osc_rdma_priority
|
|
0
|
|
errmgr
|
errmgr
|
Default selection set of components for the errmgr framework (none means "use all components that can be found")
|
none
|
|
|
errmgr_hnp_debug
|
|
0
|
|
|
errmgr_hnp_priority
|
|
0
|
|
|
errmgr_orted_debug
|
|
0
|
|
|
errmgr_orted_priority
|
|
0
|
|
|
errmgr_proxy_debug
|
|
0
|
|
|
errmgr_proxy_priority
|
|
0
|
|
gpr
|
gpr
|
Default selection set of components for the gpr framework (none means "use all components that can be found").
|
none
|
|
|
gpr_base_maxsize
|
|
2147483647
|
|
|
gpr_base_blocksize
|
|
512
|
|
|
gpr_null_priority
|
|
0
|
|
|
gpr_proxy_debug
|
|
0
|
|
|
gpr_proxy_priority
|
|
0
|
|
|
gpr_replica_debug
|
|
0
|
|
|
gpr_replica_isolate
|
|
0
|
|
|
gpr_replica_priority
|
|
0
|
|
iof
|
iof
|
Default selection set of components for the iof framework (none means "use all components that can be found").
|
none
|
|
|
iof_base_window_size
|
|
4096
|
|
|
iof_base_service
|
|
0.0.0
|
|
|
iof_proxy_debug
|
|
1
|
|
|
iof_proxy_priority
|
|
0
|
|
|
iof_svc_debug
|
|
1
|
|
|
iof_svc_priority
|
|
0
|
|
ns
|
ns
|
Default selection set of components for the ns framework (none means "use all components that can be found").
|
none
|
|
|
ns_proxy_debug
|
|
0
|
|
|
ns_proxy_maxsize
|
|
2147483647
|
|
|
ns_proxy_blocksize
|
|
512
|
|
|
ns_proxy_priority
|
|
0
|
|
|
ns_replica_debug
|
|
0
|
|
|
ns_replica_isolate
|
|
0
|
|
|
ns_replica_maxsize
|
|
2147483647
|
|
|
ns_replica_blocksize
|
|
512
|
|
|
ns_replica_priority
|
|
0
|
|
oob
|
oob
|
Default selection set of components for the oob framework (none means "use all components that can be found")
|
none
|
|
|
oob_base_verbose
|
Verbosity level for the oob framework (0 = no verbosity)
|
0
|
|
|
oob_tcp_peer_limit
|
|
-1
|
|
|
oob_tcp_peer_retries
|
|
60
|
|
|
oob_tcp_debug
|
|
0
|
|
|
oob_tcp_include
|
|
none
|
|
|
oob_tcp_exclude
|
|
none
|
|
|
oob_tcp_sndbuf
|
|
131072
|
|
|
oob_tcp_rcvbuf
|
|
131072
|
|
|
oob_tcp_connect_timeout
|
connect() timeout in seconds, before trying next interface
|
10
|
|
|
oob_tcp_connect_sleep
|
Enable (1) /Disable (0) random sleep for connection wireup
|
1
|
|
|
oob_tcp_listen_mode
|
Mode for HNP to accept incoming connections. Values are: event, listen_thread
|
event
|
|
|
oob_tcp_listen_thread_max_
queue
|
High water mark for queued accepted socket list size
|
10
|
|
|
oob_tcp_listen_thread_max_time
|
Maximum amount of time (in milliseconds) to wait between processing accepted socket list
|
10
|
|
|
oob_tcp_accept_spin_count
|
Number of times to let accept return EWOULDBLOCK before
updating accepted socket list
|
10
|
|
|
oob_tcp_priority
|
|
0
|
|
ras
|
ras
|
Default selection set of components for the ras framework (none means "use all components that can be found")
|
none
|
|
|
ras_dash_host_priority
|
Selection priority for the dash_host RAS component
|
5
|
|
|
ras_gridengine_debug
|
Enable debugging output for the gridengine RAS component
|
0
|
|
|
ras_gridengine_priority
|
Priority of the gridengine ras component
|
100
|
|
|
ras_gridengine_verbose
|
Enable verbose output for the gridengine ras component
|
0
|
|
|
ras_gridengine_show_jobid
|
Show the JOB_ID of the Grid Engine job
|
0
|
|
|
ras_hostfile_priority
|
Selection priority for the hostfile RAS component
|
10
|
|
|
ras_localhost_priority
|
Selection priority for the localhost RAS component
|
0
|
|
|
ras_tm_priority
|
Priority of the tm ras component
|
100
|
|
rds
|
rds
|
Default selection set of components for the rds framework (none means "use all components that can be found")
|
none
|
|
|
rds_hostfile_debug
|
Toggle debug output for hostfile RDS component. 1=enable; 0=disable.
|
0
|
|
|
rds_hostfile_path
|
Path to the ORTE Host filename.
|
/opt/SUNWhpc/HPC8.1/etc/openmpi-default-hostfile
|
|
|
rds_hostfile_priority
|
|
0
|
|
|
rds_proxy_priority
|
|
0
|
|
|
rds_resfile_debug
|
Toggle debug output for resfile RDS component. 1=enable; 0=disable.
|
0
|
|
|
rds_resfile_name
|
ORTE Resource filename
|
none
|
|
|
rds_resfile_priority
|
|
0
|
|
rmaps
|
rmaps
|
Default selection set of components for the rmaps framework (none means "use all components that can be found")
|
none
|
|
|
rmaps_base_verbose
|
Verbosity level for the rmaps framework
|
0
|
|
|
rmaps_base_schedule_policy
|
Scheduling policy for RMAPS. [slot | node]
|
slot
|
|
|
rmaps_base_pernode
|
Launch one ppn as directed
|
0
|
|
|
rmaps_base_n_pernode
|
Launch n procs/node
|
-1
|
|
|
rmaps_base_schedule_local
|
If nonzero, allow scheduling MPI applications on the same node as mpirun (default). If zero, do not schedule any MPI applications on the same node as mpirun.
|
1
|
|
|
rmaps_base_no_oversubscribe
|
If nonzero, then do not allow oversubscription of nodes. mpirun will return an error if there are not enough nodes to launch all processes without oversubscribing.
|
0
|
|
|
rmaps_round_robin_debug
|
Toggle debug output for Round Robin RMAPS component
|
1
|
|
|
rmaps_round_robin_priority
|
Selection priority for Round Robin RMAPS component
|
1
|
|
rmgr
|
rmgr
|
Default selection set of components for the rmgr framework (none means "use all components that can be found")
|
none
|
|
|
rmgr_proxy_priority
|
|
0
|
|
|
rmgr_urm_priority
|
|
0
|
|
rml
|
rml
|
Default selection set of components for the rml framework (none means "use all components that can be found")
|
none
|
|
|
rml_base_verbose
|
Verbosity level for the rml framework (0 = no verbosity)
|
0
|
|
|
rml_oob_priority
|
|
0
|
|
plm
|
plm
|
Default selection set of components for the plm framework (none means "use all components that can be found")
|
none
|
|
|
plm_base_reuse_daemons
|
If nonzero, reuse daemons to launch dynamically spawned processes. If zero, do not reuse daemons (default)
|
0
|
|
|
plm_fork_reap
|
Specifies whether or not to wait to reap all children before finalizing. 1=wait; 0= do not wait
|
1
|
|
|
plm_fork_reap_timeout
|
When killing child processes, first send a SIGTERM, then wait at least this timeout (in seconds), then send a SIGKILL
|
0
|
|
|
plm_fork_priority
|
Priority of this component
|
1
|
|
|
plm_fork_debug
|
Specifies whether or not to enable debugging output. 1=enable; 0= do not enable
|
0
|
|
|
plm_gridengine_debug
|
Enable debugging of gridengine plm component. 1=enable; 0= do not enable
|
0
|
|
|
plm_gridengine_verbose
|
Enable verbose output of the gridengine qrsh
-inherit command 1=enable; 0= do not enable
|
0
|
|
|
plm_gridengine_priority
|
Priority of the gridengine plm component
|
100
|
|
|
plm_gridengine_orted
|
The command name that the gridengine plm component will invoke for the ORTE daemon
|
orted
|
|
|
plm_proxy_priority
|
|
0
|
|
|
plm_rsh_debug
|
Enable debugging output for the rsh plm component 1=enable; 0= do not enable
|
0
|
|
|
plm_rsh_num_concurrent
|
How many plm_rsh_agent instances to invoke concurrently (value must be greater than 0)
|
128
|
|
|
plm_rsh_orted
|
The command name that the rsh plm component will invoke for the ORTE daemon
|
orted
|
|
|
plm_rsh_force_rsh
|
Force the launcher to always use rsh, even for local daemons
|
0
|
|
|
plm_rsh_priority
|
Priority of the rsh plm component
|
10
|
|
|
plm_rsh_delay
|
Delay (in seconds) between invocations of the remote agent, but only used when the debug MCA parameter is set to true, or when the top-level MCA debugging is enabled (otherwise this value is ignored)
|
1
|
|
|
plm_rsh_reap
|
If set to 1, wait for all the processes to complete before exiting. Otherwise, quit immediately without waiting for confirmation that all other processes in the job have completed.
|
1
|
|
|
plm_rsh_assume_same_shell
|
If set to 1, assume that the shell on the remote node is the same as the shell on the local node. Otherwise, probe for what the remote shell.
|
1
|
|
|
plm_rsh_agent
|
The command used to launch executables on remote nodes (typically either "ssh" or "rsh")
|
ssh : rsh
|
|
|
plm_tm_debug
|
Enable debugging of the TM plm. 1= enable; 0= do not enable
|
0
|
|
|
plm_tm_verbose
|
Enable verbose output of the TM plm. 1= enable; 0= do not enable
|
0
|
|
|
plm_tm_priority
|
Default selection priority
|
75
|
|
|
plm_tm_orted
|
Command to use to start proxy orted
|
orted
|
|
|
plm_tm_want_path_check
|
Specify whether the launching process should check for the plm_tm_orted executable in the PATH before launching. Any nonzero value enables this check. Note: the TM API does not give an indication of failure.
|
1
|
|
sds
|
sds
|
Default selection set of components for the sds framework (none means "use all components that can be found")
|
none
|
|
|
sds_base_verbose
|
Verbosity level for the sds framework (0 = no verbosity)
|
0
|
|
|
sds_env_priority
|
|
0
|
|
|
sds_pipe_priority
|
|
0
|
|
|
sds_seed_priority
|
|
0
|
|
|
sds_singleton_priority
|
|
0
|
|
soh
|
soh
|
Default selection set of components for the soh framework (none means "use all components that can be found")
|
none
|
