Skip to content

Incubator Pegasus
Incubator Pegasus

apache / Incubator Pegasus

通知 9
Star 5
Fork 0
Incubator Pegasus
Incubator Pegasus

前往新版Gitcode，体验更适合开发者的 AI 搜索 >>

未验证 提交 d31c1431 编写于 作者: Q QinZuoyan 提交者： GitHub
浏览文件
操作

collector: do not send email when email address is not provided; start... 

collector: do not send email when email address is not provided; start collector in onebox; improve config to support collector (#70)
上级 5e2a47ee
隐藏空白更改
内联 并排
Showing with 231 addition and 155 deletion
rdsn @ 14d8ba6e
比较 343e2e72...14d8ba6e
Subproject commit 343e2e72e4a7110e39882a93fe066fc4b6feb513
Subproject commit 14d8ba6e0c444cf4694631b70d3087a1ddda213e
run.sh
浏览文件 @ d31c1431
......@@ -437,7 +437,7 @@ function run_start_onebox()
echo "ERROR: file ${SERVER_PATH}/pegasus_server not exist"
exit -1
fi
if ps -ef | grep ' /pegasus_server config.ini' | grep -E 'app_list meta|app_list replica'; then
if ps -ef | grep ' /pegasus_server config.ini' | grep -E 'app_list meta|app_list replica|app_list collector'; then
echo "ERROR: some onebox processes are running, start failed"
exit -1
fi
......@@ -504,6 +504,15 @@ function run_start_onebox()
ps -ef | grep '/pegasus_server config.ini' | grep "\<$PID\>"
cd ..
done
mkdir -p collector
cd collector
ln -s -f ${SERVER_PATH}/pegasus_server pegasus_server
sed "s/@META_PORT@/34600/;s/@REPLICA_PORT@/34800/" ${ROOT}/config-server.ini >config.ini
echo "cd `pwd` && $PWD/pegasus_server config.ini -app_list collector &>result &"
$PWD/pegasus_server config.ini -app_list collector &>result &
PID=$!
ps -ef | grep '/pegasus_server config.ini' | grep "\<$PID\>"
cd ..
}
#####################
......@@ -533,7 +542,7 @@ function run_stop_onebox()
esac
shift
done
ps -ef | grep '/pegasus_server config.ini' | grep -E 'app_list meta|app_list replica' | awk '{print $2}' | xargs kill &>/dev/null
ps -ef | grep '/pegasus_server config.ini' | grep -E 'app_list meta|app_list replica|app_list collector' | awk '{print $2}' | xargs kill &>/dev/null
}
#####################
......@@ -563,7 +572,7 @@ function run_list_onebox()
esac
shift
done
ps -ef | grep '/pegasus_server config.ini' | grep -E 'app_list meta|app_list replica' | sort -k11
ps -ef | grep '/pegasus_server config.ini' | grep -E 'app_list meta|app_list replica|app_list collector' | sort -k11
}
#####################
......@@ -604,17 +613,17 @@ function run_clear_onebox()
function usage_start_onebox_instance()
{
echo "Options for subcommand 'start_onebox_instance':"
echo " -h|--help print the help info"
echo " -m|--meta_id <num>"
echo " meta server id"
echo " -r|--replica_id <num>"
echo " replica server id"
echo " -h|--help print the help info"
echo " -m|--meta_id <num> start meta server with id"
echo " -r|--replica_id <num> start replica server with id"
echo " -c|--collector start collector server"
}
function run_start_onebox_instance()
{
META_ID=0
REPLICA_ID=0
COLLECTOR_ID=0
while [[ $# > 0 ]]; do
key="$1"
case $key in
......@@ -630,6 +639,9 @@ function run_start_onebox_instance()
REPLICA_ID="$2"
shift
;;
-c|--collector)
COLLECTOR_ID=1
;;
*)
echo "ERROR: unknown option \"$key\""
echo
......@@ -639,12 +651,8 @@ function run_start_onebox_instance()
esac
shift
done
if [ $META_ID = "0" -a $REPLICA_ID = "0" ]; then
echo "ERROR: no meta_id or replica_id set"
exit -1
fi
if [ $META_ID != "0" -a $REPLICA_ID != "0" ]; then
echo "ERROR: meta_id and replica_id can only set one"
if [ $META_ID = "0" -a $REPLICA_ID = "0" -a $COLLECTOR_ID = "0" ]; then
echo "ERROR: no meta_id or replica_id or collector set"
exit -1
fi
if [ $META_ID != "0" ]; then
......@@ -683,6 +691,24 @@ function run_start_onebox_instance()
cd ..
echo "INFO: replica@$REPLICA_ID started"
fi
if [ $COLLECTOR_ID != "0" ]; then
dir=onebox/collector
if [ ! -d $dir ]; then
echo "ERROR: collector dir $dir not exist"
exit -1
fi
if ps -ef | grep "/collector/pegasus_server config.ini" | grep "app_list collector" ; then
echo "INFO: collector already running"
exit -1
fi
cd $dir
echo "cd `pwd` && $PWD/pegasus_server config.ini -app_list collector &>result &"
$PWD/pegasus_server config.ini -app_list collector &>result &
PID=$!
ps -ef | grep '/pegasus_server config.ini' | grep "\<$PID\>"
cd ..
echo "INFO: collector started"
fi
}
#####################
......@@ -691,17 +717,17 @@ function run_start_onebox_instance()
function usage_stop_onebox_instance()
{
echo "Options for subcommand 'stop_onebox_instance':"
echo " -h|--help print the help info"
echo " -m|--meta_id <num>"
echo " meta server id"
echo " -r|--replica_id <num>"
echo " replica server id"
echo " -h|--help print the help info"
echo " -m|--meta_id <num> stop meta server with id"
echo " -r|--replica_id <num> stop replica server with id"
echo " -c|--collector stop collector server"
}
function run_stop_onebox_instance()
{
META_ID=0
REPLICA_ID=0
COLLECTOR_ID=0
while [[ $# > 0 ]]; do
key="$1"
case $key in
......@@ -717,6 +743,9 @@ function run_stop_onebox_instance()
REPLICA_ID="$2"
shift
;;
-c|--collector)
COLLECTOR_ID=1
;;
*)
echo "ERROR: unknown option \"$key\""
echo
......@@ -726,12 +755,8 @@ function run_stop_onebox_instance()
esac
shift
done
if [ $META_ID = "0" -a $REPLICA_ID = "0" ]; then
echo "ERROR: no meta_id or replica_id set"
exit -1
fi
if [ $META_ID != "0" -a $REPLICA_ID != "0" ]; then
echo "ERROR: meta_id and replica_id can only set one"
if [ $META_ID = "0" -a $REPLICA_ID = "0" -a $COLLECTOR_ID = "0" ]; then
echo "ERROR: no meta_id or replica_id or collector set"
exit -1
fi
if [ $META_ID != "0" ]; then
......@@ -760,6 +785,14 @@ function run_stop_onebox_instance()
ps -ef | grep "/replica$REPLICA_ID/pegasus_server config.ini" | grep "app_list replica" | awk '{print $2}' | xargs kill &>/dev/null
echo "INFO: replica@$REPLICA_ID stopped"
fi
if [ $COLLECTOR_ID != "0" ]; then
if ! ps -ef | grep "/collector/pegasus_server config.ini" | grep "app_list collector" ; then
echo "INFO: collector is not running"
exit -1
fi
ps -ef | grep "/collector/pegasus_server config.ini" | grep "app_list collector" | awk '{print $2}' | xargs kill &>/dev/null
echo "INFO: collector stopped"
fi
}
#####################
......@@ -768,19 +801,18 @@ function run_stop_onebox_instance()
function usage_restart_onebox_instance()
{
echo "Options for subcommand 'restart_onebox_instance':"
echo " -h|--help print the help info"
echo " -m|--meta_id <num>"
echo " meta server id"
echo " -r|--replica_id <num>"
echo " replica server id"
echo " -s|--sleep <num>"
echo " sleep time in seconds between stop and start, default is 1"
echo " -h|--help print the help info"
echo " -m|--meta_id <num> restart meta server with id"
echo " -r|--replica_id <num> restart replica server with id"
echo " -c|--collector restart collector server"
echo " -s|--sleep <num> sleep time in seconds between stop and start, default is 1"
}
function run_restart_onebox_instance()
{
META_ID=0
REPLICA_ID=0
COLLECTOR_OPT=""
SLEEP=1
while [[ $# > 0 ]]; do
key="$1"
......@@ -797,6 +829,10 @@ function run_restart_onebox_instance()
REPLICA_ID="$2"
shift
;;
-c|--collector)
COLLECTOR_OPT="-c"
shift
;;
-s|--sleep)
SLEEP="$2"
shift
......@@ -810,18 +846,14 @@ function run_restart_onebox_instance()
esac
shift
done
if [ $META_ID = "0" -a $REPLICA_ID = "0" ]; then
echo "ERROR: no meta_id or replica_id set"
exit -1
fi
if [ $META_ID != "0" -a $REPLICA_ID != "0" ]; then
echo "ERROR: meta_id and replica_id can only set one"
if [ $META_ID = "0" -a $REPLICA_ID = "0" -a "x$COLLECTOR_OPT" = "x" ]; then
echo "ERROR: no meta_id or replica_id or collector set"
exit -1
fi
run_stop_onebox_instance -m $META_ID -r $REPLICA_ID
run_stop_onebox_instance -m $META_ID -r $REPLICA_ID $COLLECTOR_OPT
echo "sleep $SLEEP"
sleep $SLEEP
run_start_onebox_instance -m $META_ID -r $REPLICA_ID
run_start_onebox_instance -m $META_ID -r $REPLICA_ID $COLLECTOR_OPT
}
#####################
......
src/server/available_detector.cpp
浏览文件 @ d31c1431
......@@ -36,11 +36,11 @@ available_detector::available_detector()
".",
"available detect alert script dir");
dassert(_alert_script_dir.size() > 0, "");
_alert_email_address = dsn_config_get_value_string("pegasus.collector",
"available_detect_alert_email_address",
"",
"available detect alert email address");
dassert(_alert_email_address.size() > 0, "");
_alert_email_address = dsn_config_get_value_string(
"pegasus.collector",
"available_detect_alert_email_address",
"",
"available detect alert email address, empty means not send email");
_meta_list.clear();
dsn::replication::replica_helper::load_meta_servers(_meta_list);
dassert(_meta_list.size() > 0, "");
......@@ -59,17 +59,20 @@ available_detector::available_detector()
1000, // unit is millisecond,default is 1s = 1000ms
"available detect timeout");
// initialize the _client.
if (!pegasus_client_factory::initialize(nullptr))
if (!pegasus_client_factory::initialize(nullptr)) {
dassert(false, "Initialize the pegasus client failed");
}
_client = pegasus_client_factory::get_client(_cluster_name.c_str(), _app_name.c_str());
dassert(_client != nullptr, "Initialize the _client failed");
_ddl_client.reset(new replication_ddl_client(_meta_list));
dassert(_ddl_client != nullptr, "Initialize the _ddl_client failed");
_send_alert_email_cmd = "cd " + _alert_script_dir + "; bash sendmail.sh alert " +
_alert_email_address + " " + _cluster_name + " " + _app_name + " ";
_send_availability_info_email_cmd = "cd " + _alert_script_dir +
"; bash sendmail.sh availability_info " +
_alert_email_address + " " + _cluster_name + " ";
if (!_alert_email_address.empty()) {
_send_alert_email_cmd = "cd " + _alert_script_dir + "; bash sendmail.sh alert " +
_alert_email_address + " " + _cluster_name + " " + _app_name + " ";
_send_availability_info_email_cmd = "cd " + _alert_script_dir +
"; bash sendmail.sh availability_info " +
_alert_email_address + " " + _cluster_name + " ";
}
_pfc_detect_times_day.init_app_counter("app.pegasus",
"cluster.available.detect.times.day",
......@@ -130,7 +133,7 @@ void available_detector::start()
nullptr,
std::move(std::bind(&available_detector::detect_available, this)),
0,
std::chrono::seconds(60));
std::chrono::minutes(1));
report_availability_info();
}
......@@ -151,7 +154,13 @@ void available_detector::stop()
void available_detector::detect_available()
{
if (!generate_hash_keys()) {
dassert(false, "Initialize hash_keys failed and can't detect the available");
derror("initialize hash_keys failed, do not detect available, retry after 60 seconds");
_detect_timer = ::dsn::tasking::enqueue(
LPC_DETECT_AVAILABLE,
nullptr,
std::move(std::bind(&available_detector::detect_available, this)),
0,
std::chrono::minutes(1));
return;
}
_detect_tasks.clear();
......@@ -259,12 +268,13 @@ bool available_detector::generate_hash_keys()
void available_detector::on_detect(int32_t idx)
{
if (idx == 0) {
ddebug("detecting table[%s] with id[%d] on the cluster[%s], "
ddebug("detecting table[%s] with app_id[%d] and partition_count[%d] on cluster[%s], "
"recent_day_detect_times(%" PRId64 "), recent_day_fail_times(%" PRId64 "), "
"recent_hour_detect_times(%" PRId64 "), recent_hour_fail_times(%" PRId64 ") "
"recent_minute_detect_times(%" PRId64 "), recent_minute_fail_times(%" PRId64 ")",
_app_name.c_str(),
_app_id,
_partition_count,
_cluster_name.c_str(),
_recent_day_detect_times.load(),
_recent_day_fail_times.load(),
......@@ -284,50 +294,30 @@ void available_detector::on_detect(int32_t idx)
_recent_day_detect_times.fetch_add(1);
_recent_hour_detect_times.fetch_add(1);
_recent_minute_detect_times.fetch_add(1);
// define async_get callback function.
auto async_get_callback = [this, idx](
int err, std::string &&_value, pegasus_client::internal_info &&info) {
std::atomic<int> &cnt = (*_fail_count[idx]);
if (err != PERR_OK) {
int prev = cnt.fetch_add(1);
_recent_day_fail_times.fetch_add(1);
_recent_hour_fail_times.fetch_add(1);
_recent_minute_fail_times.fetch_add(1);
derror("async_get partition[%d] fail, fail_count = %d, hash_key = %s, error = %s",
idx,
prev + 1,
_hash_keys[idx].c_str(),
_client->get_error_string(err));
// check and send email.
bool send_email = false;
if (cnt.load() >= _alert_fail_count) {
::dsn::utils::auto_lock<::dsn::utils::ex_lock_nr> l(_alert_lock);
if (cnt.load() >= _alert_fail_count) {
for (auto i = 0; i < _partition_count; i++) {
std::atomic<int> &c = (*_fail_count[i]);
c.store(0);
}
send_email = true;
}
}
if (send_email) {
ddebug("start to send alert email, partition_index = %d", idx);
int r = system((_send_alert_email_cmd + std::to_string(idx)).c_str());
if (r == 0) {
ddebug("send alert email done, partition_index = %d", idx);
} else {
derror(
"send alert email fail, partition_index = %d, command_return = %d", idx, r);
}
auto async_get_callback =
[this, idx](int err, std::string &&_value, pegasus_client::internal_info &&info) {
std::atomic<int> &cnt = (*_fail_count[idx]);
if (err != PERR_OK) {
int prev = cnt.fetch_add(1);
_recent_day_fail_times.fetch_add(1);
_recent_hour_fail_times.fetch_add(1);
_recent_minute_fail_times.fetch_add(1);
derror("async_get partition[%d] fail, fail_count = %d, hash_key = %s, error = %s",
idx,
prev + 1,
_hash_keys[idx].c_str(),
_client->get_error_string(err));
check_and_send_email(&cnt, idx);
} else {
cnt.store(0);
dinfo("async_get partition[%d] ok, hash_key = %s, value = %s",
idx,
_hash_keys[idx].c_str(),
_value.c_str());
}
} else {
cnt.store(0);
dinfo("async_get partition[%d] ok, hash_key = %s, value = %s",
idx,
_hash_keys[idx].c_str(),
_value.c_str());
}
};
};
// define async_set callback function.
auto async_set_callback =
......@@ -345,28 +335,7 @@ void available_detector::on_detect(int32_t idx)
prev + 1,
_hash_keys[idx].c_str(),
_client->get_error_string(err));
// check and send email.
bool send_email = false;
if (cnt.load() >= _alert_fail_count) {
::dsn::utils::auto_lock<::dsn::utils::ex_lock_nr> l(_alert_lock);
if (cnt.load() >= _alert_fail_count) {
for (auto i = 0; i < _partition_count; i++) {
std::atomic<int> &c = (*_fail_count[i]);
c.store(0);
}
send_email = true;
}
}
if (send_email) {
ddebug("start to send alert email, partition_index = %d", idx);
int r = system((_send_alert_email_cmd + std::to_string(idx)).c_str());
if (r == 0) {
ddebug("send alert email done, partition_index = %d", idx);
} else {
derror(
"send alert email fail, partition_index = %d, command_return = %d", idx, r);
}
}
check_and_send_email(&cnt, idx);
} else {
dinfo("async_set partition[%d] ok, hash_key = %s", idx, _hash_keys[idx].c_str());
_client->async_get(
......@@ -377,16 +346,49 @@ void available_detector::on_detect(int32_t idx)
_client->async_set(_hash_keys[idx], "", value, std::move(async_set_callback), _detect_timeout);
}
void available_detector::check_and_send_email(std::atomic<int> *cnt, int32_t idx)
{
bool send_email = false;
if (cnt->load() >= _alert_fail_count) {
::dsn::utils::auto_lock<::dsn::utils::ex_lock_nr> l(_alert_lock);
if (cnt->load() >= _alert_fail_count) {
for (auto i = 0; i < _partition_count; i++) {
std::atomic<int> &c = (*_fail_count[i]);
c.store(0);
}
send_email = true;
}
}
if (send_email) {
ddebug("start to send alert email, partition_index = %d", idx);
if (_send_alert_email_cmd.empty()) {
ddebug("ignore sending alert email because email address is not set, "
"partition_index = %d",
idx);
} else {
int r = system((_send_alert_email_cmd + std::to_string(idx)).c_str());
if (r == 0) {
ddebug("send alert email done, partition_index = %d", idx);
} else {
derror("send alert email failed, partition_index = %d, "
"command_return = %d",
idx,
r);
}
}
}
}
void available_detector::on_day_report()
{
ddebug("start to report on new day, last_day = %s", _old_day.c_str());
int64_t detect_times = _recent_day_detect_times.fetch_and(0);
int64_t fail_times = _recent_day_fail_times.fetch_and(0);
int64_t succ_times = std::max(0L, detect_times - fail_times);
int64_t available = 1000000;
int64_t available = 0;
std::string hash_key("detect_available_day");
std::string sort_key(_old_day);
std::string value("0,0,1000000");
std::string value("0,0,0");
if (detect_times > 0) {
available = (int64_t)((double)succ_times / detect_times * 1000000);
std::ostringstream oss;
......@@ -399,25 +401,35 @@ void available_detector::on_day_report()
_pfc_available_day->set(available);
ddebug("start to send availability email, date = %s", _old_day.c_str());
int r = system((_send_availability_info_email_cmd + std::to_string(detect_times) + " " +
std::to_string(fail_times) + " " + _old_day)
.c_str());
if (r == 0) {
ddebug("send availability email done, date = %s, total_detect_times = %u, total_fail_times "
"= %u",
if (_send_availability_info_email_cmd.empty()) {
ddebug("ignore sending availability email because email address is not set, "
"date = %s, total_detect_times = %u, total_fail_times = %u",
_old_day.c_str(),
detect_times,
fail_times);
} else {
derror("send availability email fail, date = %s, total_detect_times = %u, total_fail_times "
"= %u, command_return = %d",
_old_day.c_str(),
detect_times,
fail_times,
r);
int r = system((_send_availability_info_email_cmd + std::to_string(detect_times) + " " +
std::to_string(fail_times) + " " + _old_day)
.c_str());
if (r == 0) {
ddebug("send availability email done, date = %s, "
"total_detect_times = %u, total_fail_times = %u",
_old_day.c_str(),
detect_times,
fail_times);
} else {
derror("send availability email fail, date = %s, "
"total_detect_times = %u, total_fail_times = %u, command_return = %d",
_old_day.c_str(),
detect_times,
fail_times,
r);
}
}
set_detect_result(hash_key, sort_key, value, 3);
// set try_count to 3000 (keep on retrying for 3000 minutes) to avoid losting detect result
// if the result table is also unavailable for a long time.
set_detect_result(hash_key, sort_key, value, 3000);
}
void available_detector::on_hour_report()
......@@ -426,10 +438,10 @@ void available_detector::on_hour_report()
int64_t detect_times = _recent_hour_detect_times.fetch_and(0);
int64_t fail_times = _recent_hour_fail_times.fetch_and(0);
int64_t succ_times = std::max(0L, detect_times - fail_times);
int64_t available = 1000000;
int64_t available = 0;
std::string hash_key("detect_available_hour");
std::string sort_key(_old_hour);
std::string value("0,0,1000000");
std::string value("0,0,0");
if (detect_times > 0) {
available = (int64_t)((double)succ_times / detect_times * 1000000);
std::ostringstream oss;
......@@ -441,7 +453,9 @@ void available_detector::on_hour_report()
_pfc_fail_times_hour->set(fail_times);
_pfc_available_hour->set(available);
set_detect_result(hash_key, sort_key, value, 3);
// set try_count to 3000 (keep on retrying for 3000 minutes) to avoid losting detect result
// if the result table is also unavailable for a long time.
set_detect_result(hash_key, sort_key, value, 3000);
}
void available_detector::on_minute_report()
......@@ -450,10 +464,10 @@ void available_detector::on_minute_report()
int64_t detect_times = _recent_minute_detect_times.fetch_and(0);
int64_t fail_times = _recent_minute_fail_times.fetch_and(0);
int64_t succ_times = std::max(0L, detect_times - fail_times);
int64_t available = 1000000;
int64_t available = 0;
std::string hash_key("detect_available_minute");
std::string sort_key(_old_minute);
std::string value("0,0,1000000");
std::string value("0,0,0");
if (detect_times > 0) {
available = (int64_t)((double)succ_times / detect_times * 1000000);
std::ostringstream oss;
......@@ -465,7 +479,9 @@ void available_detector::on_minute_report()
_pfc_fail_times_minute->set(fail_times);
_pfc_available_minute->set(available);
set_detect_result(hash_key, sort_key, value, 3);
// set try_count to 3000 (keep on retrying for 3000 minutes) to avoid losting detect result
// if the result table is also unavailable for a long time.
set_detect_result(hash_key, sort_key, value, 3000);
}
void available_detector::set_detect_result(const std::string &hash_key,
......@@ -483,7 +499,7 @@ void available_detector::set_detect_result(const std::string &hash_key,
int new_try_count = try_count - 1;
if (new_try_count > 0) {
derror("set_detect_result fail, hash_key = %s, sort_key = %s, value = %s, "
"error = %s, try_count = %d, try again after 1 minute",
"error = %s, left_try_count = %d, try again after 1 minute",
hash_key.c_str(),
sort_key.c_str(),
value.c_str(),
......@@ -499,7 +515,7 @@ void available_detector::set_detect_result(const std::string &hash_key,
std::chrono::minutes(1));
} else {
derror("set_detect_result fail, hash_key = %s, sort_key = %s, value = %s, "
"error = %s, try_count = %d, do not try again",
"error = %s, left_try_count = %d, do not try again",
hash_key.c_str(),
sort_key.c_str(),
value.c_str(),
......
src/server/available_detector.h
浏览文件 @ d31c1431
......@@ -32,6 +32,7 @@ private:
// generate hash_keys that can access every partition.
bool generate_hash_keys();
void on_detect(int32_t idx);
void check_and_send_email(std::atomic<int> *cnt, int32_t idx);
void detect_available();
void report_availability_info();
......
src/server/config-server.ini
浏览文件 @ d31c1431
......@@ -23,6 +23,15 @@ pools = THREAD_POOL_DEFAULT,THREAD_POOL_REPLICATION_LONG,THREAD_POOL_REPLICATION
run = true
count = 1
[apps.collector]
name = collector
type = collector
arguments =
ports = 34100
pools = THREAD_POOL_DEFAULT,THREAD_POOL_REPLICATION
run = true
count = 1
[core]
;tool = simulator
;tool = fastrun
......@@ -157,7 +166,7 @@ perf_test_concurrent = true
perf_test_concurrent_count = 20
[meta_server]
server_list = @LOCAL_IP@:34601, @LOCAL_IP@:34602, @LOCAL_IP@:34603
server_list = @LOCAL_IP@:34601,@LOCAL_IP@:34602,@LOCAL_IP@:34603
cluster_root = /pegasus/onebox/@LOCAL_IP@
meta_state_service_type = meta_state_service_zookeeper
meta_state_service_parameters =
......@@ -187,7 +196,7 @@ stateful = true
package_id =
[replication]
data_dirs_black_list_file = /home/mi/.pegasus_data_dirs_black_list
data_dirs_black_list_file = ./.pegasus_data_dirs_black_list
deny_client_on_start = false
delay_for_fd_timeout_on_start = false
......@@ -261,6 +270,20 @@ falcon_host = 127.0.0.1
falcon_port = 1988
falcon_path = /v1/push
[pegasus.collector]
cluster = onebox
available_detect_app = @APP_NAME@
available_detect_alert_script_dir = ./package/bin
available_detect_alert_email_address =
available_detect_interval_seconds = 3
available_detect_alert_fail_count = 30
available_detect_timeout = 5000
app_stat_interval_seconds = 10
[uri-resolver.dsn://onebox]
factory = partition_resolver_simple
arguments = @LOCAL_IP@:34601,@LOCAL_IP@:34602,@LOCAL_IP@:34603
[components.pegasus_perf_counter_number_percentile_atomic]
counter_computation_interval_seconds = 10
......
src/server/config.ini
浏览文件 @ d31c1431
......@@ -20,6 +20,15 @@
run = true
count = 1
[apps.collector]
name = collector
type = collector
arguments =
ports = 34101
pools = THREAD_POOL_DEFAULT,THREAD_POOL_REPLICATION
run = true
count = 1
[apps.mimic]
name = mimic
type = dsn.app.mimic
......@@ -29,15 +38,6 @@
count = 1
delay_seconds = 30
[apps.collector]
name = collector
type = collector
arguments =
ports = 34101
pools = THREAD_POOL_DEFAULT, THREAD_POOL_REPLICATION
run = true
count = 1
[core]
data_dir = %{app.dir}
......@@ -80,7 +80,7 @@
max_file_copy_request_count_per_file = 10
[network]
primary_interface = %{network.interface}
primary_interface =
io_service_worker_count = 4
[threadpool..default]
......@@ -291,12 +291,16 @@
cluster = %{cluster.name}
available_detect_app = temp
available_detect_alert_script_dir = ./package/bin
available_detect_alert_email_address = %{email.address}
available_detect_alert_email_address =
available_detect_interval_seconds = 3
available_detect_alert_fail_count = 30
available_detect_timeout = 5000
app_stat_interval_seconds = 10
[uri-resolver.dsn://%{cluster.name}]
factory = partition_resolver_simple
arguments = %{meta.server.list}
[components.pegasus_perf_counter_number_percentile_atomic]
counter_computation_interval_seconds = 10
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或