13th, 14th question are as follows:
(13) Some messages are not shown even though --mca snapc_base_verbose parameter
is used.
Framework : snapc
Component : full
The source file : orte/mca/snapc/base/snapc_base_open.c
The function name : orte_snapc_base_open
I think that the following verbose messages are not shown.
Because the orte_snapc_base_output ID is not initialized at those point.
The orte_snapc_base_output ID is initialized in opal_output_set_verbosity
function called by mca_base_components_open function.
OPAL_OUTPUT_VERBOSE((10, orte_snapc_base_output,
"snapc:base: open()"));
OPAL_OUTPUT_VERBOSE((20, orte_snapc_base_output,
"snapc:base: open: base_global_snapshot_dir = %s",
orte_snapc_base_global_snapshot_dir));
OPAL_OUTPUT_VERBOSE((20, orte_snapc_base_output,
"snapc:base: open: base_store_in_place = %d",
orte_snapc_base_store_in_place));
OPAL_OUTPUT_VERBOSE((20, orte_snapc_base_output,
"snapc:base: open: base_only_one_seq = %d",
orte_snapc_base_store_only_one_seq));
OPAL_OUTPUT_VERBOSE((20, orte_snapc_base_output,
"snapc:base: open: base_establish_global_snapshot_dir
= %d",
orte_snapc_base_establish_global_snapshot_dir));
OPAL_OUTPUT_VERBOSE((20, orte_snapc_base_output,
"snapc:base: open: base_global_snapshot_ref = %s",
orte_snapc_base_global_snapshot_ref));
Result of running.
previous messages are not shown.
mca: base: components_open: Looking for snapc components
mca: base: components_open: opening snapc components
mca: base: components_open: found loaded component full
mca: base: components_open: component full has no register function
snapc:full: open()
snapc:full: open: priority = 20
snapc:full: open: verbosity = 100
snapc:full: open: skip_filem = False
mca: base: components_open: component full open function successful
snapc:select: Using none component
snapc:full: close()
(14) I use the Aggregate MCA parameter -am ft-enable-cr to enable
checkpoint/restart
fault tolerance for an MPI application.
If two or more mca parameter files are specified by -am option,
Fault tolerance may be disabled.
I understand it is a specification of Open MPI.
Is there any way to specify multiple AMCA parameter files?
For user,it will be convenient to use MCA parameter file.
For example:
mpiexec .... --mca btl self,tcp -am ft-enable-cr -am
/home/guest/Test/CR-Debug/local-mca-param.conf .... a.out
-bash-3.2$ cat local-mca-param.conf
crs_base_snapshot_dir=/home/guest/Test/temp
snapc_base_global_snapshot_dir=/home/guest/Test/cr-work
-bash-3.2$
Fault tolerance is disabled.
