1
0
Fork 0

Bug fix 3.3: Fix supervisor thread crash (#4165)

* port devel branch to 3.3 of supervisor thread death fix
This commit is contained in:
Matthew Von-Maszewski 2017-12-27 16:34:29 -05:00 committed by Max Neunhöffer
parent ef8fcd101c
commit 392ddde251
1 changed files with 170 additions and 153 deletions

View File

@ -120,19 +120,19 @@ struct HealthRecord {
obj.add("Host", VPackValue(hostId));
obj.add("SyncStatus", VPackValue(syncStatus));
obj.add("Status", VPackValue(status));
if (syncTime.empty()) {
if (syncTime.empty()) {
obj.add("Timestamp",
VPackValue(timepointToString(std::chrono::system_clock::now())));
VPackValue(timepointToString(std::chrono::system_clock::now())));
} else {
obj.add("SyncTime", VPackValue(syncTime));
obj.add("LastAcked", VPackValue(lastAcked));
}
}
bool statusDiff(HealthRecord const& other) {
return (status != other.status || syncStatus != other.syncStatus);
}
friend std::ostream& operator<<(std::ostream& o, HealthRecord const& hr) {
VPackBuilder builder;
{ VPackObjectBuilder b(&builder);
@ -140,7 +140,7 @@ struct HealthRecord {
o << builder.toJson();
return o;
}
};
@ -193,7 +193,7 @@ void Supervision::upgradeOne(Builder& builder) {
builder.add("oldEmpty", VPackValue(true));
}
}
}
}
}
}
@ -221,7 +221,7 @@ void Supervision::upgradeHealthRecords(Builder& builder) {
// "/arango/Supervision/health" is in old format
Builder b;
size_t n = 0;
if (_snapshot.has(healthPrefix)) {
HealthRecord hr;
{ VPackObjectBuilder oo(&b);
@ -234,14 +234,14 @@ void Supervision::upgradeHealthRecords(Builder& builder) {
b.add(VPackValue(recPair.first));
{ VPackObjectBuilder ooo(&b);
hr.toVelocyPack(b);
}
}
}
}
}
}
if (n>0) {
{ VPackArrayBuilder trx(&builder);
{ VPackObjectBuilder o(&builder);
@ -249,7 +249,7 @@ void Supervision::upgradeHealthRecords(Builder& builder) {
}
}
}
}
// Upgrade agency, guarded by wakeUp
@ -280,10 +280,10 @@ void handleOnStatusDBServer(
Agent* agent, Node const& snapshot, HealthRecord& persisted,
HealthRecord& transisted, std::string const& serverID,
uint64_t const& jobId, std::shared_ptr<VPackBuilder>& envelope) {
std::string failedServerPath = failedServersPrefix + "/" + serverID;
// New condition GOOD:
// New condition GOOD:
if (transisted.status == Supervision::HEALTH_STATUS_GOOD) {
if (snapshot.has(failedServerPath)) {
Builder del;
@ -307,14 +307,14 @@ void handleOnStatusDBServer(
"supervision", serverID).create(envelope);
}
}
}
void handleOnStatusCoordinator(
Agent* agent, Node const& snapshot, HealthRecord& persisted,
HealthRecord& transisted, std::string const& serverID) {
if (transisted.status == Supervision::HEALTH_STATUS_FAILED) {
// if the current foxxmaster server failed => reset the value to ""
if (snapshot.has(foxxmaster) && snapshot(foxxmaster).getString() == serverID) {
@ -324,7 +324,7 @@ void handleOnStatusCoordinator(
create.add(foxxmaster, VPackValue("")); }}
singleWriteTransaction(agent, create);
}
}
}
@ -333,7 +333,7 @@ void handleOnStatusSingle(
HealthRecord& transisted, std::string const& serverID) {
// if the current leader server failed => reset the value to ""
if (transisted.status == Supervision::HEALTH_STATUS_FAILED) {
if (snapshot.has(asyncReplLeader) && snapshot(asyncReplLeader).getString() == serverID) {
VPackBuilder create;
{ VPackArrayBuilder tx(&create);
@ -341,10 +341,10 @@ void handleOnStatusSingle(
create.add(asyncReplLeader, VPackValue("")); }}
singleWriteTransaction(agent, create);
}
}
}
void handleOnStatus(
Agent* agent, Node const& snapshot, HealthRecord& persisted,
HealthRecord& transisted, std::string const& serverID,
@ -362,7 +362,7 @@ void handleOnStatus(
LOG_TOPIC(ERR, Logger::SUPERVISION)
<< "Unknown server type. No supervision action taken. " << serverID;
}
}
// Check all DB servers, guarded above doChecks
@ -397,7 +397,7 @@ std::vector<check_t> Supervision::check(std::string const& type) {
for (auto const& srv : todelete) {
{ VPackObjectBuilder server(del.get());
del->add(VPackValue(_agencyPrefix + healthPrefix + srv));
{ VPackObjectBuilder oper(del.get());
{ VPackObjectBuilder oper(del.get());
del->add("op", VPackValue("delete")); }}}}}
_agent->write(del);
}
@ -405,122 +405,131 @@ std::vector<check_t> Supervision::check(std::string const& type) {
// Do actual monitoring
for (auto const& machine : machinesPlanned) {
std::string lastHeartbeatStatus, lastHeartbeatAcked, lastHeartbeatTime,
lastStatus, serverID(machine.first),
shortName(_snapshot(targetShortID + serverID + "/ShortName").getString());
// Endpoint
std::string endpoint;
std::string epPath = serverID + "/endpoint";
if (serversRegistered.has(epPath)) {
endpoint = serversRegistered(epPath).getString();
}
std::string hostId;
std::string hoPath = serverID + "/host";
if (serversRegistered.has(hoPath)) {
hostId = serversRegistered(hoPath).getString();
}
lastStatus, serverID(machine.first), shortName;
// Health records from persistence, from transience and a new one
HealthRecord transist(shortName, endpoint, hostId);
HealthRecord persist(shortName, endpoint, hostId);
// Get last health entries from transient and persistent key value stores
if (_transient.has(healthPrefix + serverID)) {
transist = _transient(healthPrefix + serverID);
}
if (_snapshot.has(healthPrefix + serverID)) {
persist = _snapshot(healthPrefix + serverID);
}
// New health record (start with old add current information from sync)
// Sync.time is copied to Health.syncTime
// Sync.status is copied to Health.syncStatus
std::string syncTime = _transient.has(syncPrefix + serverID) ?
_transient(syncPrefix + serverID + "/time").getString() :
timepointToString(std::chrono::system_clock::time_point());
std::string syncStatus = _transient.has(syncPrefix + serverID) ?
_transient(syncPrefix + serverID + "/status").getString() : "UNKNOWN";
// short name arrives asynchronous to machine registering, make sure
// it has arrived before trying to use it
if (LEAF == _snapshot(targetShortID + serverID + "/ShortName").type()) {
// Last change registered in sync (transient != sync)
// Either now or value in transient
auto lastAckedTime = (syncTime != transist.syncTime) ?
std::chrono::system_clock::now() : stringToTimepoint(transist.lastAcked);
transist.lastAcked = timepointToString(lastAckedTime);
transist.syncTime = syncTime;
transist.syncStatus = syncStatus;
// Calculate elapsed since lastAcked
auto elapsed = std::chrono::duration<double>(
std::chrono::system_clock::now() - lastAckedTime);
if (elapsed.count() <= _okThreshold) {
transist.status = Supervision::HEALTH_STATUS_GOOD;
} else if (elapsed.count() <= _gracePeriod) {
transist.status = Supervision::HEALTH_STATUS_BAD;
} else {
transist.status = Supervision::HEALTH_STATUS_FAILED;
}
shortName = _snapshot(targetShortID + serverID + "/ShortName").getString();
// Status changed?
bool changed = transist.statusDiff(persist);
// Endpoint
std::string endpoint;
std::string epPath = serverID + "/endpoint";
if (serversRegistered.has(epPath)) {
endpoint = serversRegistered(epPath).getString();
}
std::string hostId;
std::string hoPath = serverID + "/host";
if (serversRegistered.has(hoPath)) {
hostId = serversRegistered(hoPath).getString();
}
// Take necessary actions if any
std::shared_ptr<VPackBuilder> envelope;
if (changed) {
handleOnStatus(_agent, _snapshot, persist, transist, serverID, _jobId,
envelope);
}
persist = transist; // Now copy Status, SyncStatus from transient to persited
// Transient report
std::shared_ptr<Builder> tReport = std::make_shared<Builder>();
{ VPackArrayBuilder transaction(tReport.get()); // Transist Transaction
// Health records from persistence, from transience and a new one
HealthRecord transist(shortName, endpoint, hostId);
HealthRecord persist(shortName, endpoint, hostId);
// Get last health entries from transient and persistent key value stores
if (_transient.has(healthPrefix + serverID)) {
transist = _transient(healthPrefix + serverID);
}
if (_snapshot.has(healthPrefix + serverID)) {
persist = _snapshot(healthPrefix + serverID);
}
// New health record (start with old add current information from sync)
// Sync.time is copied to Health.syncTime
// Sync.status is copied to Health.syncStatus
std::string syncTime = _transient.has(syncPrefix + serverID) ?
_transient(syncPrefix + serverID + "/time").getString() :
timepointToString(std::chrono::system_clock::time_point());
std::string syncStatus = _transient.has(syncPrefix + serverID) ?
_transient(syncPrefix + serverID + "/status").getString() : "UNKNOWN";
// Last change registered in sync (transient != sync)
// Either now or value in transient
auto lastAckedTime = (syncTime != transist.syncTime) ?
std::chrono::system_clock::now() : stringToTimepoint(transist.lastAcked);
transist.lastAcked = timepointToString(lastAckedTime);
transist.syncTime = syncTime;
transist.syncStatus = syncStatus;
// Calculate elapsed since lastAcked
auto elapsed = std::chrono::duration<double>(
std::chrono::system_clock::now() - lastAckedTime);
if (elapsed.count() <= _okThreshold) {
transist.status = Supervision::HEALTH_STATUS_GOOD;
} else if (elapsed.count() <= _gracePeriod) {
transist.status = Supervision::HEALTH_STATUS_BAD;
} else {
transist.status = Supervision::HEALTH_STATUS_FAILED;
}
// Status changed?
bool changed = transist.statusDiff(persist);
// Take necessary actions if any
std::shared_ptr<VPackBuilder> envelope;
{ VPackObjectBuilder operation(tReport.get()); // Operation
tReport->add(VPackValue(healthPrefix + serverID)); // Supervision/Health
{ VPackObjectBuilder oo(tReport.get());
transist.toVelocyPack(*tReport); }}} // Transaction
// Persistent report
std::shared_ptr<Builder> pReport = nullptr;
if (changed) {
pReport = std::make_shared<Builder>();
{ VPackArrayBuilder transaction(pReport.get()); // Persist Transaction
{ VPackObjectBuilder operation(pReport.get()); // Operation
pReport->add(VPackValue(healthPrefix + serverID)); // Supervision/Health
{ VPackObjectBuilder oo(pReport.get());
persist.toVelocyPack(*pReport); }
if (envelope != nullptr) { // Failed server
TRI_ASSERT(
envelope->slice().isArray() && envelope->slice()[0].isObject());
for (VPackObjectIterator::ObjectPair i : VPackObjectIterator(envelope->slice()[0])) {
pReport->add(i.key.copyString(), i.value);
}
}} // Operation
if (envelope != nullptr) { // Preconditions(Job)
TRI_ASSERT(
envelope->slice().isArray() && envelope->slice()[1].isObject());
pReport->add(envelope->slice()[1]);
}} // Transaction
}
if (!this->isStopping()) {
// Replicate special event and only then transient store
if (changed) {
write_ret_t res = singleWriteTransaction(_agent, *pReport);
if (res.accepted && res.indices.front() != 0) {
++_jobId; // Job was booked
handleOnStatus(_agent, _snapshot, persist, transist, serverID, _jobId,
envelope);
}
persist = transist; // Now copy Status, SyncStatus from transient to persited
// Transient report
std::shared_ptr<Builder> tReport = std::make_shared<Builder>();
{ VPackArrayBuilder transaction(tReport.get()); // Transist Transaction
std::shared_ptr<VPackBuilder> envelope;
{ VPackObjectBuilder operation(tReport.get()); // Operation
tReport->add(VPackValue(healthPrefix + serverID)); // Supervision/Health
{ VPackObjectBuilder oo(tReport.get());
transist.toVelocyPack(*tReport); }}} // Transaction
// Persistent report
std::shared_ptr<Builder> pReport = nullptr;
if (changed) {
pReport = std::make_shared<Builder>();
{ VPackArrayBuilder transaction(pReport.get()); // Persist Transaction
{ VPackObjectBuilder operation(pReport.get()); // Operation
pReport->add(VPackValue(healthPrefix + serverID)); // Supervision/Health
{ VPackObjectBuilder oo(pReport.get());
persist.toVelocyPack(*pReport); }
if (envelope != nullptr) { // Failed server
TRI_ASSERT(
envelope->slice().isArray() && envelope->slice()[0].isObject());
for (VPackObjectIterator::ObjectPair i : VPackObjectIterator(envelope->slice()[0])) {
pReport->add(i.key.copyString(), i.value);
}
}} // Operation
if (envelope != nullptr) { // Preconditions(Job)
TRI_ASSERT(
envelope->slice().isArray() && envelope->slice()[1].isObject());
pReport->add(envelope->slice()[1]);
}} // Transaction
}
if (!this->isStopping()) {
// Replicate special event and only then transient store
if (changed) {
write_ret_t res = singleWriteTransaction(_agent, *pReport);
if (res.accepted && res.indices.front() != 0) {
++_jobId; // Job was booked
transient(_agent, *tReport);
}
} else { // Nothing special just transient store
transient(_agent, *tReport);
}
} else { // Nothing special just transient store
transient(_agent, *tReport);
}
}
}
} else {
LOG_TOPIC(INFO, Logger::SUPERVISION) <<
"Short name for << " << serverID << " not yet available. Skipping health check.";
} // else
} // for
return ret;
}
@ -528,22 +537,22 @@ std::vector<check_t> Supervision::check(std::string const& type) {
// Update local agency snapshot, guarded by callers
bool Supervision::updateSnapshot() {
_lock.assertLockedByCurrentThread();
if (_agent == nullptr || this->isStopping()) {
return false;
}
_agent->executeLocked([&]() {
if (_agent->readDB().has(_agencyPrefix)) {
_snapshot = _agent->readDB().get(_agencyPrefix);
}
if (_agent->transient().has(_agencyPrefix)) {
_transient = _agent->transient().get(_agencyPrefix);
}
}
});
return true;
}
// All checks, guarded by main thread
@ -570,8 +579,8 @@ void Supervision::run() {
CONDITION_LOCKER(guard, _cv);
_cv.wait(static_cast<uint64_t>(1000000 * _frequency));
}
bool done = false;
bool done = false;
MUTEX_LOCKER(locker, _lock);
_agent->executeLocked([&]() {
if (_agent->readDB().has(supervisionNode)) {
@ -586,7 +595,7 @@ void Supervision::run() {
}
}
});
if (done) {
break;
}
@ -604,7 +613,7 @@ void Supervision::run() {
{
MUTEX_LOCKER(locker, _lock);
// Get bunch of job IDs from agency for future jobs
if (_agent->leading() && (_jobId == 0 || _jobId == _jobIdMax)) {
getUniqueIds(); // cannot fail but only hang
@ -621,9 +630,17 @@ void Supervision::run() {
// Do nothing unless leader for over 10 seconds
auto secondsSinceLeader = std::chrono::duration<double>(
std::chrono::steady_clock::now() - _agent->leaderSince()).count();
// 10 seconds should be plenty of time for all servers to send
// heartbeat status to new leader (heartbeat is once per second)
if (secondsSinceLeader > 10.0) {
doChecks();
try {
doChecks();
} catch (std::exception const& e) {
LOG_TOPIC(ERR, Logger::SUPERVISION) << e.what() << " " << __FILE__ << " " << __LINE__;
} catch (...) {
LOG_TOPIC(ERR, Logger::SUPERVISION) <<
"Supervision::doChecks() generated an uncaught exception.";
}
}
}
@ -710,11 +727,11 @@ void Supervision::handleShutdown() {
}
}
// Guarded by caller
// Guarded by caller
bool Supervision::handleJobs() {
_lock.assertLockedByCurrentThread();
// Do supervision
shrinkCluster();
enforceReplication();
workJobs();
@ -735,7 +752,7 @@ void Supervision::workJobs() {
JobContext(
PENDING, (*pendEnt.second)("jobId").getString(), _snapshot, _agent).run();
}
}
@ -747,7 +764,7 @@ void Supervision::enforceReplication() {
auto const& db = *(db_.second);
for (const auto& col_ : db.children()) { // Planned collections
auto const& col = *(col_.second);
size_t replicationFactor;
if (col.has("replicationFactor") && col("replicationFactor").isUInt()) {
replicationFactor = col("replicationFactor").getUInt();
@ -762,13 +779,13 @@ void Supervision::enforceReplication() {
auto available = Job::availableServers(_snapshot);
replicationFactor = available.size();
}
bool clone = col.has("distributeShardsLike");
if (!clone) {
for (auto const& shard_ : col("shards").children()) { // Pl shards
auto const& shard = *(shard_.second);
size_t actualReplicationFactor = shard.slice().length();
if (actualReplicationFactor != replicationFactor) {
// Check that there is not yet an addFollower or removeFollower
@ -810,7 +827,7 @@ void Supervision::enforceReplication() {
}
}
}
}
void Supervision::fixPrototypeChain(Builder& migrate) {
@ -827,7 +844,7 @@ void Supervision::fixPrototypeChain(Builder& migrate) {
}
return (s.empty()) ? col : resolve(db, s);
};
for (auto const& database : _snapshot(planColPrefix).children()) {
for (auto const& collection : database.second->children()) {
if (collection.second->has("distributeShardsLike")) {
@ -865,13 +882,13 @@ void Supervision::shrinkCluster() {
if (!todo.empty() || !pending.empty()) { // This is low priority
return;
}
// Get servers from plan
auto availServers = Job::availableServers(_snapshot);
size_t targetNumDBServers;
std::string const NDBServers ("/Target/NumberOfDBServers");
if (_snapshot.has(NDBServers) && _snapshot(NDBServers).isUInt()) {
targetNumDBServers = _snapshot(NDBServers).getUInt();
} else {
@ -932,7 +949,7 @@ void Supervision::shrinkCluster() {
availServers.size() > targetNumDBServers) {
// Sort servers by name
std::sort(availServers.begin(), availServers.end());
// Schedule last server for cleanout
CleanOutServer(_snapshot, _agent, std::to_string(_jobId++),
"supervision", availServers.back()).run();
@ -978,7 +995,7 @@ void Supervision::getUniqueIds() {
{ VPackArrayBuilder a(builder.get());
builder->add(VPackValue(path)); }
} // [[{path:{"op":"increment","step":n}}],[path]]
auto ret = _agent->transact(builder);
if (ret.accepted) {
try {
@ -989,10 +1006,10 @@ void Supervision::getUniqueIds() {
} catch (std::exception const& e) {
LOG_TOPIC(ERR, Logger::SUPERVISION)
<< "Failed to acquire job IDs from agency: "
<< e.what() << __FILE__ << __LINE__;
<< e.what() << __FILE__ << " " << __LINE__;
}
}
}