a200e4f865
*** THIS RFC INCLUDES A MINOR CHANGE TO THE MPI-RTE INTERFACE *** Note: during the course of this work, it was necessary to completely separate the MPI and RTE progress engines. There were multiple places in the MPI layer where ORTE_WAIT_FOR_COMPLETION was being used. A new OMPI_WAIT_FOR_COMPLETION macro was created (defined in ompi/mca/rte/rte.h) that simply cycles across opal_progress until the provided flag becomes false. Places where the MPI layer blocked waiting for RTE to complete an event have been modified to use this macro. *************************************************************************************** I am reissuing this RFC because of the time that has passed since its original release. Since its initial release and review, I have debugged it further to ensure it fully supports tests like loop_spawn. It therefore seems ready for merge back to the trunk. Given its prior review, I have set the timeout for one week. The code is in https://bitbucket.org/rhc/ompi-oob2 WHAT: Rewrite of ORTE OOB WHY: Support asynchronous progress and a host of other features WHEN: Wed, August 21 SYNOPSIS: The current OOB has served us well, but a number of limitations have been identified over the years. Specifically: * it is only progressed when called via opal_progress, which can lead to hangs or recursive calls into libevent (which is not supported by that code) * we've had issues when multiple NICs are available as the code doesn't "shift" messages between transports - thus, all nodes had to be available via the same TCP interface. * the OOB "unloads" incoming opal_buffer_t objects during the transmission, thus preventing use of OBJ_RETAIN in the code when repeatedly sending the same message to multiple recipients * there is no failover mechanism across NICs - if the selected NIC (or its attached switch) fails, we are forced to abort * only one transport (i.e., component) can be "active" The revised OOB resolves these problems: * async progress is used for all application processes, with the progress thread blocking in the event library * each available TCP NIC is supported by its own TCP module. The ability to asynchronously progress each module independently is provided, but not enabled by default (a runtime MCA parameter turns it "on") * multi-address TCP NICs (e.g., a NIC with both an IPv4 and IPv6 address, or with virtual interfaces) are supported - reachability is determined by comparing the contact info for a peer against all addresses within the range covered by the address/mask pairs for the NIC. * a message that arrives on one TCP NIC is automatically shifted to whatever NIC that is connected to the next "hop" if that peer cannot be reached by the incoming NIC. If no TCP module will reach the peer, then the OOB attempts to send the message via all other available components - if none can reach the peer, then an "error" is reported back to the RML, which then calls the errmgr for instructions. * opal_buffer_t now conforms to standard object rules re OBJ_RETAIN as we no longer "unload" the incoming object * NIC failure is reported to the TCP component, which then tries to resend the message across any other available TCP NIC. If that doesn't work, then the message is given back to the OOB base to try using other components. If all that fails, then the error is reported to the RML, which reports to the errmgr for instructions * obviously from the above, multiple OOB components (e.g., TCP and UD) can be active in parallel * the matching code has been moved to the RML (and out of the OOB/TCP component) so it is independent of transport * routing is done by the individual OOB modules (as opposed to the RML). Thus, both routed and non-routed transports can simultaneously be active * all blocking send/recv APIs have been removed. Everything operates asynchronously. KNOWN LIMITATIONS: * although provision is made for component failover as described above, the code for doing so has not been fully implemented yet. At the moment, if all connections for a given peer fail, the errmgr is notified of a "lost connection", which by default results in termination of the job if it was a lifeline * the IPv6 code is present and compiles, but is not complete. Since the current IPv6 support in the OOB doesn't work anyway, I don't consider this a blocker * routing is performed at the individual module level, yet the active routed component is selected on a global basis. We probably should update that to reflect that different transports may need/choose to route in different ways * obviously, not every error path has been tested nor necessarily covered * determining abnormal termination is more challenging than in the old code as we now potentially have multiple ways of connecting to a process. Ideally, we would declare "connection failed" when *all* transports can no longer reach the process, but that requires some additional (possibly complex) code. For now, the code replicates the old behavior only somewhat modified - i.e., if a module sees its connection fail, it checks to see if it is a lifeline. If so, it notifies the errmgr that the lifeline is lost - otherwise, it notifies the errmgr that a non-lifeline connection was lost. * reachability is determined solely on the basis of a shared subnet address/mask - more sophisticated algorithms (e.g., the one used in the tcp btl) are required to handle routing via gateways * the RML needs to assign sequence numbers to each message on a per-peer basis. The receiving RML will then deliver messages in order, thus preventing out-of-order messaging in the case where messages travel across different transports or a message needs to be redirected/resent due to failure of a NIC This commit was SVN r29058.
281 строка
11 KiB
C
281 строка
11 KiB
C
/*
|
|
* Copyright (c) 2011 Los Alamos National Security, LLC.
|
|
* All rights reserved.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
/**** ORTE STATE MACHINE ****/
|
|
|
|
/* States are treated as events so that the event
|
|
* library can sequence them. Each state consists
|
|
* of an event, a job or process state, a pointer
|
|
* to the respective object, and a callback function
|
|
* to be executed for that state. Events can be defined
|
|
* at different priorities - e.g., SYS priority for
|
|
* events associated with launching jobs, and ERR priority
|
|
* for events associated with abnormal termination of
|
|
* a process.
|
|
*
|
|
* The state machine consists of a list of state objects,
|
|
* each defining a state-cbfunc pair. At startup, a default
|
|
* list is created by the base functions which is then
|
|
* potentially customized by selected components within
|
|
* the various ORTE frameworks. For example, a PLM component
|
|
* may need to insert states in the launch procedure, or may
|
|
* want to redirect a particular state callback to a custom
|
|
* function.
|
|
*
|
|
* For convenience, an ANY state can be defined along with a generic
|
|
* callback function, with the corresponding state object
|
|
* placed at the end of the state machine. Setting the
|
|
* machine to a state that has not been explicitly defined
|
|
* will cause this default action to be executed. Thus, you
|
|
* don't have to explicitly define a state-cbfunc pair
|
|
* for every job or process state.
|
|
*/
|
|
|
|
#ifndef _ORTE_STATE_H_
|
|
#define _ORTE_STATE_H_
|
|
|
|
#include "orte_config.h"
|
|
|
|
#include "opal/class/opal_list.h"
|
|
#include "opal/mca/event/event.h"
|
|
|
|
#include "orte/mca/plm/plm_types.h"
|
|
#include "orte/runtime/orte_globals.h"
|
|
|
|
#include "orte/mca/state/state_types.h"
|
|
|
|
BEGIN_C_DECLS
|
|
|
|
/*
|
|
* MCA Framework - put here to access the opal_output channel
|
|
* in the macros
|
|
*/
|
|
ORTE_DECLSPEC extern mca_base_framework_t orte_state_base_framework;
|
|
|
|
/* For ease in debugging the state machine, it is STRONGLY recommended
|
|
* that the functions be accessed using the following macros
|
|
*/
|
|
#define ORTE_FORCED_TERMINATE(x) \
|
|
do { \
|
|
if (!orte_abnormal_term_ordered) { \
|
|
opal_output_verbose(1, orte_state_base_framework.framework_output, \
|
|
"%s FORCE-TERMINATE AT %s:%d", \
|
|
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME), \
|
|
__FILE__, __LINE__); \
|
|
ORTE_UPDATE_EXIT_STATUS(x); \
|
|
ORTE_ACTIVATE_JOB_STATE(NULL, ORTE_JOB_STATE_FORCED_EXIT); \
|
|
/* set the global abnormal exit flag */ \
|
|
orte_abnormal_term_ordered = true; \
|
|
} \
|
|
} while(0);
|
|
|
|
#define ORTE_ACTIVATE_JOB_STATE(j, s) \
|
|
do { \
|
|
orte_job_t *shadow=(j); \
|
|
opal_output_verbose(1, orte_state_base_framework.framework_output, \
|
|
"%s ACTIVATE JOB %s STATE %s AT %s:%d", \
|
|
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME), \
|
|
(NULL == shadow) ? "NULL" : \
|
|
ORTE_JOBID_PRINT(shadow->jobid), \
|
|
orte_job_state_to_str((s)), \
|
|
__FILE__, __LINE__); \
|
|
orte_state.activate_job_state(shadow, (s)); \
|
|
} while(0);
|
|
|
|
#define ORTE_ACTIVATE_PROC_STATE(p, s) \
|
|
do { \
|
|
orte_process_name_t *shadow=(p); \
|
|
opal_output_verbose(1, orte_state_base_framework.framework_output, \
|
|
"%s ACTIVATE PROC %s STATE %s AT %s:%d", \
|
|
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME), \
|
|
(NULL == shadow) ? "NULL" : \
|
|
ORTE_NAME_PRINT(shadow), \
|
|
orte_proc_state_to_str((s)), \
|
|
__FILE__, __LINE__); \
|
|
orte_state.activate_proc_state(shadow, (s)); \
|
|
} while(0);
|
|
|
|
/**
|
|
* Module initialization function.
|
|
*
|
|
* @retval ORTE_SUCCESS The operation completed successfully
|
|
* @retval ORTE_ERROR An unspecifed error occurred
|
|
*/
|
|
typedef int (*orte_state_base_module_init_fn_t)(void);
|
|
|
|
/**
|
|
* Module finalization function.
|
|
*
|
|
* @retval ORTE_SUCCESS The operation completed successfully
|
|
* @retval ORTE_ERROR An unspecifed error occurred
|
|
*/
|
|
typedef int (*orte_state_base_module_finalize_fn_t)(void);
|
|
|
|
/**** JOB STATE APIs ****/
|
|
/* Job states are accessed via orte_job_t objects as they are only
|
|
* used in ORTE tools and not application processes. APIs are provided
|
|
* for assembling and editing the state machine, as well as activating
|
|
* a specific job state
|
|
*
|
|
* Note the inherent assumption in this design that any customization
|
|
* of the state machine will at least start with the base states - i.e.,
|
|
* that one would start with the default machine and edit it to add,
|
|
* remove, or modify callbacks as required. Alternatively, one could
|
|
* just clear the list entirely and assemble a fully custom state
|
|
* machine - both models are supported.
|
|
*/
|
|
|
|
/* Activate a state in the job state machine.
|
|
*
|
|
* Creates and activates an event with the callback corresponding to the
|
|
* specified job state. If the specified state is not found:
|
|
*
|
|
* 1. if a state machine entry for ORTE_JOB_STATE_ERROR was given, and
|
|
* the state is an error state (i.e., ORTE_JOB_STATE_ERROR <= state),
|
|
* then the callback for the ERROR state will be used
|
|
*
|
|
* 2. if a state machine entry for ORTE_JOB_STATE_ANY was given, and
|
|
* the state is not an error state (i.e., state < ORTE_JOB_STATE_ERROR),
|
|
* then the callback for the ANY state will be used
|
|
*
|
|
* 3. if neither of the above is true, then the call will be ignored.
|
|
*/
|
|
typedef void (*orte_state_base_module_activate_job_state_fn_t)(orte_job_t *jdata,
|
|
orte_job_state_t state);
|
|
|
|
/* Add a state to the job state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_add_job_state_fn_t)(orte_job_state_t state,
|
|
orte_state_cbfunc_t cbfunc,
|
|
int priority);
|
|
|
|
/* Set the callback function for a state in the job state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_set_job_state_callback_fn_t)(orte_job_state_t state,
|
|
orte_state_cbfunc_t cbfunc);
|
|
|
|
/* Set the event priority for a state in the job state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_set_job_state_priority_fn_t)(orte_job_state_t state,
|
|
int priority);
|
|
|
|
/* Remove a state from the job state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_remove_job_state_fn_t)(orte_job_state_t state);
|
|
|
|
|
|
/**** Proc STATE APIs ****/
|
|
/* Proc states are accessed via orte_process_name_t as the state machine
|
|
* must be available to both application processes and ORTE tools. APIs are
|
|
* providedfor assembling and editing the state machine, as well as activating
|
|
* a specific proc state
|
|
*
|
|
* Note the inherent assumption in this design that any customization
|
|
* of the state machine will at least start with the base states - i.e.,
|
|
* that one would start with the default machine and edit it to add,
|
|
* remove, or modify callbacks as required. Alternatively, one could
|
|
* just clear the list entirely and assemble a fully custom state
|
|
* machine - both models are supported.
|
|
*/
|
|
|
|
/* Activate a proc state.
|
|
*
|
|
* Creates and activates an event with the callback corresponding to the
|
|
* specified proc state. If the specified state is not found:
|
|
*
|
|
* 1. if a state machine entry for ORTE_PROC_STATE_ERROR was given, and
|
|
* the state is an error state (i.e., ORTE_PROC_STATE_ERROR <= state),
|
|
* then the callback for the ERROR state will be used
|
|
*
|
|
* 2. if a state machine entry for ORTE_PROC_STATE_ANY was given, and
|
|
* the state is not an error state (i.e., state < ORTE_PROC_STATE_ERROR),
|
|
* then the callback for the ANY state will be used
|
|
*
|
|
* 3. if neither of the above is true, then the call will be ignored.
|
|
*/
|
|
typedef void (*orte_state_base_module_activate_proc_state_fn_t)(orte_process_name_t *proc,
|
|
orte_proc_state_t state);
|
|
|
|
/* Add a state to the proc state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_add_proc_state_fn_t)(orte_proc_state_t state,
|
|
orte_state_cbfunc_t cbfunc,
|
|
int priority);
|
|
|
|
/* Set the callback function for a state in the proc state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_set_proc_state_callback_fn_t)(orte_proc_state_t state,
|
|
orte_state_cbfunc_t cbfunc);
|
|
|
|
/* Set the event priority for a state in the proc state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_set_proc_state_priority_fn_t)(orte_proc_state_t state,
|
|
int priority);
|
|
|
|
/* Remove a state from the proc state machine.
|
|
*
|
|
*/
|
|
typedef int (*orte_state_base_module_remove_proc_state_fn_t)(orte_proc_state_t state);
|
|
|
|
|
|
/*
|
|
* Module Structure
|
|
*/
|
|
struct orte_state_base_module_1_0_0_t {
|
|
/** Initialization Function */
|
|
orte_state_base_module_init_fn_t init;
|
|
/** Finalization Function */
|
|
orte_state_base_module_finalize_fn_t finalize;
|
|
/* Job state APIs */
|
|
orte_state_base_module_activate_job_state_fn_t activate_job_state;
|
|
orte_state_base_module_add_job_state_fn_t add_job_state;
|
|
orte_state_base_module_set_job_state_callback_fn_t set_job_state_callback;
|
|
orte_state_base_module_set_job_state_priority_fn_t set_job_state_priority;
|
|
orte_state_base_module_remove_job_state_fn_t remove_job_state;
|
|
/* Proc state APIs */
|
|
orte_state_base_module_activate_proc_state_fn_t activate_proc_state;
|
|
orte_state_base_module_add_proc_state_fn_t add_proc_state;
|
|
orte_state_base_module_set_proc_state_callback_fn_t set_proc_state_callback;
|
|
orte_state_base_module_set_proc_state_priority_fn_t set_proc_state_priority;
|
|
orte_state_base_module_remove_proc_state_fn_t remove_proc_state;
|
|
};
|
|
typedef struct orte_state_base_module_1_0_0_t orte_state_base_module_1_0_0_t;
|
|
typedef orte_state_base_module_1_0_0_t orte_state_base_module_t;
|
|
ORTE_DECLSPEC extern orte_state_base_module_t orte_state;
|
|
|
|
/*
|
|
* State Component
|
|
*/
|
|
struct orte_state_base_component_1_0_0_t {
|
|
/** MCA base component */
|
|
mca_base_component_t base_version;
|
|
/** MCA base data */
|
|
mca_base_component_data_t base_data;
|
|
};
|
|
typedef struct orte_state_base_component_1_0_0_t orte_state_base_component_1_0_0_t;
|
|
typedef orte_state_base_component_1_0_0_t orte_state_base_component_t;
|
|
|
|
/*
|
|
* Macro for use in components that are of type state
|
|
*/
|
|
#define ORTE_STATE_BASE_VERSION_1_0_0 \
|
|
MCA_BASE_VERSION_2_0_0, \
|
|
"state", 1, 0, 0
|
|
|
|
END_C_DECLS
|
|
#endif
|