boinc/sched/message_handler.cpp

188 lines
5.0 KiB
C++
Raw Normal View History

// This file is part of BOINC.
// http://boinc.berkeley.edu
// Copyright (C) 2019 University of California
//
// BOINC is free software; you can redistribute it and/or modify it
// under the terms of the GNU Lesser General Public License
// as published by the Free Software Foundation,
// either version 3 of the License, or (at your option) any later version.
//
// BOINC is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
// See the GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with BOINC. If not, see <http://www.gnu.org/licenses/>.
// message_handler - check and validate new messages
- server: change the following from per-host to per-(host, app version): - daily quota mechanism - reliable mechanism (accelerated retries) - "trusted" mechanism (adaptive replication) - scheduler: enforce host scale probation only for apps with host_scale_check set. - validator: do scale probation on invalid results (need this in addition to error and timeout cases) - feeder: update app version scales every 10 min, not 10 sec - back-end apps: support --foo as well as -foo for options Notes: - If you have, say, cuda, cuda23 and cuda_fermi plan classes, a host will have separate quotas for each one. That means it could error out on 100 jobs for cuda_fermi, and when its quota goes to zero, error out on 100 jobs for cuda23, etc. This is intentional; there may be cases where one version works but not the others. - host.error_rate and host.max_results_day are deprecated TODO: - the values in the app table for limits on jobs in progress etc. should override rather than config.xml. Implementation notes: scheduler: process_request(): read all host_app_versions for host at start; Compute "reliable" and "trusted" for each one. write modified records at end get_app_version(): add "reliable_only" arg; if set, use only reliable versions skip over-quota versions Multi-pass scheduling: if have at least one reliable version, do a pass for jobs that need reliable, and use only reliable versions. Then clear best_app_versions cache. Score-based scheduling: for need-reliable jobs, it will pick the fastest version, then give a score bonus if that version happens to be reliable. When get back a successful result from client: increase daily quota When get back an error result from client: impose scale probation decrease daily quota if not aborted Validator: when handling a WU, create a vector of HOST_APP_VERSION parallel to vector of RESULT. Pass it to assign_credit_set(). Make copies of originals so we can update only modified ones update HOST_APP_VERSION error rates Transitioner: decrease quota on timeout svn path=/trunk/boinc/; revision=21181
2010-04-15 03:13:56 +00:00
// [--d debug_level]
// [--one_pass] // make one pass through table, then exit
//
// int handle_message(MSG_FROM_HOST&)
// handle a message from the host
//
// return nonzero on error
#include "config.h"
#include <unistd.h>
#include <cstdlib>
#include <string>
#include "boinc_db.h"
#include "util.h"
#include "error_numbers.h"
#include "str_replace.h"
#include "str_util.h"
#include "svn_version.h"
#include "sched_config.h"
#include "sched_util.h"
#include "sched_msgs.h"
char app_name[256];
extern int handle_message(MSG_FROM_HOST&);
int handle_message(MSG_FROM_HOST& mfh) {
int retval;
printf("got message \n%s\n", mfh.xml);
DB_MSG_TO_HOST mth;
mth.clear();
mth.create_time = time(0);
mth.hostid = mfh.hostid;
mth.handled = false;
safe_strcpy(mth.xml, mfh.xml);
retval = mth.insert();
if (retval) {
printf("insert failed %s\n", boincerror(retval));
}
return 0;
}
// make one pass through msgs_from_host with handled == 0
// return true if there were any
//
bool do_message_scan() {
DB_MSG_FROM_HOST mfh;
char buf[256];
bool found=false;
int retval;
sprintf(buf, "where handled=0");
while (1) {
retval = mfh.enumerate(buf);
if (retval) {
if (retval != ERR_DB_NOT_FOUND) {
log_messages.printf(MSG_DEBUG,
"DB connection lost, exiting\n"
);
exit(0);
}
break;
}
retval = handle_message(mfh);
if (!retval) {
mfh.handled = true;
mfh.update();
}
found = true;
}
return found;
}
int main_loop(bool one_pass) {
int retval;
bool did_something;
// char buf[256];
retval = boinc_db.open(config.db_name, config.db_host, config.db_user, config.db_passwd);
if (retval) {
log_messages.printf(MSG_CRITICAL,
"boinc_db.open failed: %s\n", boinc_db.error_string()
);
exit(1);
}
/*
sprintf(buf, "where name='%s'", app_name);
retval = app.lookup(buf);
if (retval) {
log_messages.printf(MSG_CRITICAL,
"can't find app %s\n", app_name
);
exit(1);
}
*/
// coverity[loop_top] - infinite loop is intended
while (1) {
check_stop_daemons();
did_something = do_message_scan();
if (one_pass) break;
if (!did_something) {
daemon_sleep(5);
}
}
return 0;
}
void usage(char *name) {
fprintf(stderr,
"check and validate new messages\n\n"
"Usage: %s [OPTION]...\n\n"
"Options:\n"
" [ -d X ] Set debug level to X\n"
- server: change the following from per-host to per-(host, app version): - daily quota mechanism - reliable mechanism (accelerated retries) - "trusted" mechanism (adaptive replication) - scheduler: enforce host scale probation only for apps with host_scale_check set. - validator: do scale probation on invalid results (need this in addition to error and timeout cases) - feeder: update app version scales every 10 min, not 10 sec - back-end apps: support --foo as well as -foo for options Notes: - If you have, say, cuda, cuda23 and cuda_fermi plan classes, a host will have separate quotas for each one. That means it could error out on 100 jobs for cuda_fermi, and when its quota goes to zero, error out on 100 jobs for cuda23, etc. This is intentional; there may be cases where one version works but not the others. - host.error_rate and host.max_results_day are deprecated TODO: - the values in the app table for limits on jobs in progress etc. should override rather than config.xml. Implementation notes: scheduler: process_request(): read all host_app_versions for host at start; Compute "reliable" and "trusted" for each one. write modified records at end get_app_version(): add "reliable_only" arg; if set, use only reliable versions skip over-quota versions Multi-pass scheduling: if have at least one reliable version, do a pass for jobs that need reliable, and use only reliable versions. Then clear best_app_versions cache. Score-based scheduling: for need-reliable jobs, it will pick the fastest version, then give a score bonus if that version happens to be reliable. When get back a successful result from client: increase daily quota When get back an error result from client: impose scale probation decrease daily quota if not aborted Validator: when handling a WU, create a vector of HOST_APP_VERSION parallel to vector of RESULT. Pass it to assign_credit_set(). Make copies of originals so we can update only modified ones update HOST_APP_VERSION error rates Transitioner: decrease quota on timeout svn path=/trunk/boinc/; revision=21181
2010-04-15 03:13:56 +00:00
" [ --one_pass ] make one pass through table, then exit\n"
" [ -h --help ] show this help text.\n"
" [ -v | --version ] show version information\n",
name
);
}
int main(int argc, char** argv) {
int i, retval;
2007-04-18 20:49:58 +00:00
bool one_pass = false;
check_stop_daemons();
for (i=1; i<argc; i++) {
- server: change the following from per-host to per-(host, app version): - daily quota mechanism - reliable mechanism (accelerated retries) - "trusted" mechanism (adaptive replication) - scheduler: enforce host scale probation only for apps with host_scale_check set. - validator: do scale probation on invalid results (need this in addition to error and timeout cases) - feeder: update app version scales every 10 min, not 10 sec - back-end apps: support --foo as well as -foo for options Notes: - If you have, say, cuda, cuda23 and cuda_fermi plan classes, a host will have separate quotas for each one. That means it could error out on 100 jobs for cuda_fermi, and when its quota goes to zero, error out on 100 jobs for cuda23, etc. This is intentional; there may be cases where one version works but not the others. - host.error_rate and host.max_results_day are deprecated TODO: - the values in the app table for limits on jobs in progress etc. should override rather than config.xml. Implementation notes: scheduler: process_request(): read all host_app_versions for host at start; Compute "reliable" and "trusted" for each one. write modified records at end get_app_version(): add "reliable_only" arg; if set, use only reliable versions skip over-quota versions Multi-pass scheduling: if have at least one reliable version, do a pass for jobs that need reliable, and use only reliable versions. Then clear best_app_versions cache. Score-based scheduling: for need-reliable jobs, it will pick the fastest version, then give a score bonus if that version happens to be reliable. When get back a successful result from client: increase daily quota When get back an error result from client: impose scale probation decrease daily quota if not aborted Validator: when handling a WU, create a vector of HOST_APP_VERSION parallel to vector of RESULT. Pass it to assign_credit_set(). Make copies of originals so we can update only modified ones update HOST_APP_VERSION error rates Transitioner: decrease quota on timeout svn path=/trunk/boinc/; revision=21181
2010-04-15 03:13:56 +00:00
if (is_arg(argv[i], "one_pass")) {
one_pass = true;
- server: change the following from per-host to per-(host, app version): - daily quota mechanism - reliable mechanism (accelerated retries) - "trusted" mechanism (adaptive replication) - scheduler: enforce host scale probation only for apps with host_scale_check set. - validator: do scale probation on invalid results (need this in addition to error and timeout cases) - feeder: update app version scales every 10 min, not 10 sec - back-end apps: support --foo as well as -foo for options Notes: - If you have, say, cuda, cuda23 and cuda_fermi plan classes, a host will have separate quotas for each one. That means it could error out on 100 jobs for cuda_fermi, and when its quota goes to zero, error out on 100 jobs for cuda23, etc. This is intentional; there may be cases where one version works but not the others. - host.error_rate and host.max_results_day are deprecated TODO: - the values in the app table for limits on jobs in progress etc. should override rather than config.xml. Implementation notes: scheduler: process_request(): read all host_app_versions for host at start; Compute "reliable" and "trusted" for each one. write modified records at end get_app_version(): add "reliable_only" arg; if set, use only reliable versions skip over-quota versions Multi-pass scheduling: if have at least one reliable version, do a pass for jobs that need reliable, and use only reliable versions. Then clear best_app_versions cache. Score-based scheduling: for need-reliable jobs, it will pick the fastest version, then give a score bonus if that version happens to be reliable. When get back a successful result from client: increase daily quota When get back an error result from client: impose scale probation decrease daily quota if not aborted Validator: when handling a WU, create a vector of HOST_APP_VERSION parallel to vector of RESULT. Pass it to assign_credit_set(). Make copies of originals so we can update only modified ones update HOST_APP_VERSION error rates Transitioner: decrease quota on timeout svn path=/trunk/boinc/; revision=21181
2010-04-15 03:13:56 +00:00
} else if (is_arg(argv[i], "d")) {
if (!argv[++i]) {
log_messages.printf(MSG_CRITICAL, "%s requires an argument\n\n", argv[--i]);
usage(argv[0]);
exit(1);
}
int dl = atoi(argv[i]);
log_messages.set_debug_level(dl);
if (dl == 4) g_print_queries = true;
- server: change the following from per-host to per-(host, app version): - daily quota mechanism - reliable mechanism (accelerated retries) - "trusted" mechanism (adaptive replication) - scheduler: enforce host scale probation only for apps with host_scale_check set. - validator: do scale probation on invalid results (need this in addition to error and timeout cases) - feeder: update app version scales every 10 min, not 10 sec - back-end apps: support --foo as well as -foo for options Notes: - If you have, say, cuda, cuda23 and cuda_fermi plan classes, a host will have separate quotas for each one. That means it could error out on 100 jobs for cuda_fermi, and when its quota goes to zero, error out on 100 jobs for cuda23, etc. This is intentional; there may be cases where one version works but not the others. - host.error_rate and host.max_results_day are deprecated TODO: - the values in the app table for limits on jobs in progress etc. should override rather than config.xml. Implementation notes: scheduler: process_request(): read all host_app_versions for host at start; Compute "reliable" and "trusted" for each one. write modified records at end get_app_version(): add "reliable_only" arg; if set, use only reliable versions skip over-quota versions Multi-pass scheduling: if have at least one reliable version, do a pass for jobs that need reliable, and use only reliable versions. Then clear best_app_versions cache. Score-based scheduling: for need-reliable jobs, it will pick the fastest version, then give a score bonus if that version happens to be reliable. When get back a successful result from client: increase daily quota When get back an error result from client: impose scale probation decrease daily quota if not aborted Validator: when handling a WU, create a vector of HOST_APP_VERSION parallel to vector of RESULT. Pass it to assign_credit_set(). Make copies of originals so we can update only modified ones update HOST_APP_VERSION error rates Transitioner: decrease quota on timeout svn path=/trunk/boinc/; revision=21181
2010-04-15 03:13:56 +00:00
} else if (is_arg(argv[i], "h") || is_arg(argv[i], "help")) {
usage(argv[0]);
exit(0);
- server: change the following from per-host to per-(host, app version): - daily quota mechanism - reliable mechanism (accelerated retries) - "trusted" mechanism (adaptive replication) - scheduler: enforce host scale probation only for apps with host_scale_check set. - validator: do scale probation on invalid results (need this in addition to error and timeout cases) - feeder: update app version scales every 10 min, not 10 sec - back-end apps: support --foo as well as -foo for options Notes: - If you have, say, cuda, cuda23 and cuda_fermi plan classes, a host will have separate quotas for each one. That means it could error out on 100 jobs for cuda_fermi, and when its quota goes to zero, error out on 100 jobs for cuda23, etc. This is intentional; there may be cases where one version works but not the others. - host.error_rate and host.max_results_day are deprecated TODO: - the values in the app table for limits on jobs in progress etc. should override rather than config.xml. Implementation notes: scheduler: process_request(): read all host_app_versions for host at start; Compute "reliable" and "trusted" for each one. write modified records at end get_app_version(): add "reliable_only" arg; if set, use only reliable versions skip over-quota versions Multi-pass scheduling: if have at least one reliable version, do a pass for jobs that need reliable, and use only reliable versions. Then clear best_app_versions cache. Score-based scheduling: for need-reliable jobs, it will pick the fastest version, then give a score bonus if that version happens to be reliable. When get back a successful result from client: increase daily quota When get back an error result from client: impose scale probation decrease daily quota if not aborted Validator: when handling a WU, create a vector of HOST_APP_VERSION parallel to vector of RESULT. Pass it to assign_credit_set(). Make copies of originals so we can update only modified ones update HOST_APP_VERSION error rates Transitioner: decrease quota on timeout svn path=/trunk/boinc/; revision=21181
2010-04-15 03:13:56 +00:00
} else if (is_arg(argv[i], "v") || is_arg(argv[i], "version")) {
printf("%s\n", SVN_VERSION);
exit(0);
} else {
log_messages.printf(MSG_CRITICAL, "unknown command line argument: %s\n\n", argv[i]);
usage(argv[0]);
exit(1);
}
}
retval = config.parse_file();
if (retval) {
log_messages.printf(MSG_CRITICAL,
"Can't parse config.xml: %s\n", boincerror(retval)
);
exit(1);
}
log_messages.printf(MSG_NORMAL, "Starting message handler\n");
install_stop_signal_handler();
main_loop(one_pass);
}