wbtiger
/
foundationdb
mirror of https://github.com/apple/foundationdb.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
foundationdb/fdbclient/ManagementAPI.actor.cpp

2535 lines
92 KiB
C++
Raw Blame History

/*
* ManagementAPI.actor.cpp
*
* This source file is part of the FoundationDB open source project
*
* Copyright 2013-2018 Apple Inc. and the FoundationDB project authors
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include <cinttypes>
#include <string>
#include <vector>
#include "fdbclient/Knobs.h"
#include "flow/Arena.h"
#include "fdbclient/FDBOptions.g.h"
#include "fdbclient/FDBTypes.h"
#include "fdbclient/ReadYourWrites.h"
#include "fdbclient/ManagementAPI.actor.h"
#include "fdbclient/SystemData.h"
#include "fdbclient/NativeAPI.actor.h"
#include "fdbclient/CoordinationInterface.h"
#include "fdbclient/DatabaseContext.h"
#include "fdbrpc/simulator.h"
#include "fdbclient/StatusClient.h"
#include "flow/Trace.h"
#include "flow/UnitTest.h"
#include "fdbrpc/ReplicationPolicy.h"
#include "fdbrpc/Replication.h"
#include "flow/actorcompiler.h" // This must be the last #include.
bool isInteger(const std::string& s) {
if (s.empty())
return false;
char* p;
strtol(s.c_str(), &p, 10);
return (*p == 0);
}
// Defines the mapping between configuration names (as exposed by fdbcli, buildConfiguration()) and actual configuration
// parameters
std::map<std::string, std::string> configForToken(std::string const& mode) {
std::map<std::string, std::string> out;
std::string p = configKeysPrefix.toString();
if (mode == "new") {
out[p + "initialized"] = "1";
return out;
}
if (mode == "tss") {
// Set temporary marker in config map to mark that this is a tss configuration and not a normal storage/log
// configuration. A bit of a hack but reuses the parsing code nicely.
out[p + "istss"] = "1";
return out;
}
if (mode == "locked") {
// Setting this key is interpreted as an instruction to use the normal version-stamp-based mechanism for locking
// the database.
out[databaseLockedKey.toString()] = deterministicRandom()->randomUniqueID().toString();
return out;
}
size_t pos;
// key:=value is unvalidated and unchecked
pos = mode.find(":=");
if (pos != std::string::npos) {
out[p + mode.substr(0, pos)] = mode.substr(pos + 2);
return out;
}
// key=value is constrained to a limited set of options and basic validation is performed
pos = mode.find("=");
if (pos != std::string::npos) {
std::string key = mode.substr(0, pos);
std::string value = mode.substr(pos + 1);
if (key == "proxies" && isInteger(value)) {
printf("Warning: Proxy role is being split into GRV Proxy and Commit Proxy, now prefer configuring "
"'grv_proxies' and 'commit_proxies' separately. Generally we should follow that 'commit_proxies'"
" is three times of 'grv_proxies' count and 'grv_proxies' should be not more than 4.\n");
int proxiesCount = atoi(value.c_str());
if (proxiesCount == -1) {
proxiesCount = CLIENT_KNOBS->DEFAULT_AUTO_GRV_PROXIES + CLIENT_KNOBS->DEFAULT_AUTO_COMMIT_PROXIES;
ASSERT_WE_THINK(proxiesCount >= 2);
}
if (proxiesCount < 2) {
printf("Error: At least 2 proxies (1 GRV proxy and 1 Commit proxy) are required.\n");
return out;
}
int grvProxyCount = std::max(1,
std::min(CLIENT_KNOBS->DEFAULT_MAX_GRV_PROXIES,
proxiesCount / (CLIENT_KNOBS->DEFAULT_COMMIT_GRV_PROXIES_RATIO + 1)));
int commitProxyCount = proxiesCount - grvProxyCount;
ASSERT_WE_THINK(grvProxyCount >= 1 && commitProxyCount >= 1);
out[p + "grv_proxies"] = std::to_string(grvProxyCount);
out[p + "commit_proxies"] = std::to_string(commitProxyCount);
printf("%d proxies are automatically converted into %d GRV proxies and %d Commit proxies.\n",
proxiesCount,
grvProxyCount,
commitProxyCount);
TraceEvent("DatabaseConfigurationProxiesSpecified")
.detail("SpecifiedProxies", atoi(value.c_str()))
.detail("EffectiveSpecifiedProxies", proxiesCount)
.detail("ConvertedGrvProxies", grvProxyCount)
.detail("ConvertedCommitProxies", commitProxyCount);
}
if ((key == "logs" || key == "commit_proxies" || key == "grv_proxies" || key == "resolvers" ||
key == "remote_logs" || key == "log_routers" || key == "usable_regions" ||
key == "repopulate_anti_quorum" || key == "count") &&
isInteger(value)) {
out[p + key] = value;
}
if (key == "regions") {
json_spirit::mValue mv;
json_spirit::read_string(value, mv);
StatusObject regionObj;
regionObj["regions"] = mv;
out[p + key] =
BinaryWriter::toValue(regionObj, IncludeVersion(ProtocolVersion::withRegionConfiguration())).toString();
}
if (key == "perpetual_storage_wiggle" && isInteger(value)) {
int ppWiggle = atoi(value.c_str());
if (ppWiggle >= 2 || ppWiggle < 0) {
printf("Error: Only 0 and 1 are valid values of perpetual_storage_wiggle at present.\n");
return out;
}
out[p + key] = value;
}
return out;
}
Optional<KeyValueStoreType> logType;
Optional<KeyValueStoreType> storeType;
if (mode == "ssd-1") {
logType = KeyValueStoreType::SSD_BTREE_V1;
storeType = KeyValueStoreType::SSD_BTREE_V1;
} else if (mode == "ssd" || mode == "ssd-2") {
logType = KeyValueStoreType::SSD_BTREE_V2;
storeType = KeyValueStoreType::SSD_BTREE_V2;
} else if (mode == "ssd-redwood-experimental") {
logType = KeyValueStoreType::SSD_BTREE_V2;
storeType = KeyValueStoreType::SSD_REDWOOD_V1;
} else if (mode == "ssd-rocksdb-experimental") {
logType = KeyValueStoreType::SSD_BTREE_V2;
storeType = KeyValueStoreType::SSD_ROCKSDB_V1;
} else if (mode == "memory" || mode == "memory-2") {
logType = KeyValueStoreType::SSD_BTREE_V2;
storeType = KeyValueStoreType::MEMORY;
} else if (mode == "memory-1") {
logType = KeyValueStoreType::MEMORY;
storeType = KeyValueStoreType::MEMORY;
} else if (mode == "memory-radixtree-beta") {
logType = KeyValueStoreType::SSD_BTREE_V2;
storeType = KeyValueStoreType::MEMORY_RADIXTREE;
}
// Add any new store types to fdbserver/workloads/ConfigureDatabase, too
if (storeType.present()) {
out[p + "log_engine"] = format("%d", logType.get().storeType());
out[p + "storage_engine"] = format("%d", KeyValueStoreType::StoreType(storeType.get()));
return out;
}
std::string redundancy, log_replicas;
Reference<IReplicationPolicy> storagePolicy;
Reference<IReplicationPolicy> tLogPolicy;
bool redundancySpecified = true;
if (mode == "single") {
redundancy = "1";
log_replicas = "1";
storagePolicy = tLogPolicy = Reference<IReplicationPolicy>(new PolicyOne());
} else if (mode == "double" || mode == "fast_recovery_double") {
redundancy = "2";
log_replicas = "2";
storagePolicy = tLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())));
} else if (mode == "triple" || mode == "fast_recovery_triple") {
redundancy = "3";
log_replicas = "3";
storagePolicy = tLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(3, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())));
} else if (mode == "three_datacenter" || mode == "multi_dc") {
redundancy = "6";
log_replicas = "4";
storagePolicy = Reference<IReplicationPolicy>(
new PolicyAcross(3,
"dcid",
Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())))));
tLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2,
"dcid",
Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())))));
} else if (mode == "three_datacenter_fallback") {
redundancy = "4";
log_replicas = "4";
storagePolicy = tLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2,
"dcid",
Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())))));
} else if (mode == "three_data_hall") {
redundancy = "3";
log_replicas = "4";
storagePolicy = Reference<IReplicationPolicy>(
new PolicyAcross(3, "data_hall", Reference<IReplicationPolicy>(new PolicyOne())));
tLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2,
"data_hall",
Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())))));
} else if (mode == "three_data_hall_fallback") {
redundancy = "2";
log_replicas = "4";
storagePolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2, "data_hall", Reference<IReplicationPolicy>(new PolicyOne())));
tLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2,
"data_hall",
Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())))));
} else
redundancySpecified = false;
if (redundancySpecified) {
out[p + "storage_replicas"] = redundancy;
out[p + "log_replicas"] = log_replicas;
out[p + "log_anti_quorum"] = "0";
BinaryWriter policyWriter(IncludeVersion(ProtocolVersion::withReplicationPolicy()));
serializeReplicationPolicy(policyWriter, storagePolicy);
out[p + "storage_replication_policy"] = policyWriter.toValue().toString();
policyWriter = BinaryWriter(IncludeVersion(ProtocolVersion::withReplicationPolicy()));
serializeReplicationPolicy(policyWriter, tLogPolicy);
out[p + "log_replication_policy"] = policyWriter.toValue().toString();
return out;
}
std::string remote_redundancy, remote_log_replicas;
Reference<IReplicationPolicy> remoteTLogPolicy;
bool remoteRedundancySpecified = true;
if (mode == "remote_default") {
remote_redundancy = "0";
remote_log_replicas = "0";
remoteTLogPolicy = Reference<IReplicationPolicy>();
} else if (mode == "remote_single") {
remote_redundancy = "1";
remote_log_replicas = "1";
remoteTLogPolicy = Reference<IReplicationPolicy>(new PolicyOne());
} else if (mode == "remote_double") {
remote_redundancy = "2";
remote_log_replicas = "2";
remoteTLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())));
} else if (mode == "remote_triple") {
remote_redundancy = "3";
remote_log_replicas = "3";
remoteTLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(3, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())));
} else if (mode == "remote_three_data_hall") { // FIXME: not tested in simulation
remote_redundancy = "3";
remote_log_replicas = "4";
remoteTLogPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(2,
"data_hall",
Reference<IReplicationPolicy>(
new PolicyAcross(2, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())))));
} else
remoteRedundancySpecified = false;
if (remoteRedundancySpecified) {
out[p + "remote_log_replicas"] = remote_log_replicas;
BinaryWriter policyWriter(IncludeVersion(ProtocolVersion::withReplicationPolicy()));
serializeReplicationPolicy(policyWriter, remoteTLogPolicy);
out[p + "remote_log_policy"] = policyWriter.toValue().toString();
return out;
}
return out;
}
ConfigurationResult buildConfiguration(std::vector<StringRef> const& modeTokens,
std::map<std::string, std::string>& outConf) {
for (auto it : modeTokens) {
std::string mode = it.toString();
auto m = configForToken(mode);
if (!m.size()) {
TraceEvent(SevWarnAlways, "UnknownOption").detail("Option", mode);
return ConfigurationResult::UNKNOWN_OPTION;
}
for (auto t = m.begin(); t != m.end(); ++t) {
if (outConf.count(t->first)) {
TraceEvent(SevWarnAlways, "ConflictingOption").detail("Option", t->first);
return ConfigurationResult::CONFLICTING_OPTIONS;
}
outConf[t->first] = t->second;
}
}
auto p = configKeysPrefix.toString();
if (!outConf.count(p + "storage_replication_policy") && outConf.count(p + "storage_replicas")) {
int storageCount = stoi(outConf[p + "storage_replicas"]);
Reference<IReplicationPolicy> storagePolicy = Reference<IReplicationPolicy>(
new PolicyAcross(storageCount, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())));
BinaryWriter policyWriter(IncludeVersion(ProtocolVersion::withReplicationPolicy()));
serializeReplicationPolicy(policyWriter, storagePolicy);
outConf[p + "storage_replication_policy"] = policyWriter.toValue().toString();
}
if (!outConf.count(p + "log_replication_policy") && outConf.count(p + "log_replicas")) {
int logCount = stoi(outConf[p + "log_replicas"]);
Reference<IReplicationPolicy> logPolicy = Reference<IReplicationPolicy>(
new PolicyAcross(logCount, "zoneid", Reference<IReplicationPolicy>(new PolicyOne())));
BinaryWriter policyWriter(IncludeVersion(ProtocolVersion::withReplicationPolicy()));
serializeReplicationPolicy(policyWriter, logPolicy);
outConf[p + "log_replication_policy"] = policyWriter.toValue().toString();
}
if (outConf.count(p + "istss")) {
// redo config parameters to be tss config instead of normal config
// save param values from parsing as a normal config
bool isNew = outConf.count(p + "initialized");
Optional<std::string> count;
Optional<std::string> storageEngine;
if (outConf.count(p + "count")) {
count = Optional<std::string>(outConf[p + "count"]);
}
if (outConf.count(p + "storage_engine")) {
storageEngine = Optional<std::string>(outConf[p + "storage_engine"]);
}
// A new tss setup must have count + storage engine. An adjustment must have at least one.
if ((isNew && (!count.present() || !storageEngine.present())) ||
(!isNew && !count.present() && !storageEngine.present())) {
return ConfigurationResult::INCOMPLETE_CONFIGURATION;
}
// clear map and only reset tss parameters
outConf.clear();
if (count.present()) {
outConf[p + "tss_count"] = count.get();
}
if (storageEngine.present()) {
outConf[p + "tss_storage_engine"] = storageEngine.get();
}
}
return ConfigurationResult::SUCCESS;
}
ConfigurationResult buildConfiguration(std::string const& configMode, std::map<std::string, std::string>& outConf) {
std::vector<StringRef> modes;
int p = 0;
while (p < configMode.size()) {
int end = configMode.find_first_of(' ', p);
if (end == configMode.npos)
end = configMode.size();
modes.push_back(StringRef(configMode).substr(p, end - p));
p = end + 1;
}
return buildConfiguration(modes, outConf);
}
bool isCompleteConfiguration(std::map<std::string, std::string> const& options) {
std::string p = configKeysPrefix.toString();
return options.count(p + "log_replicas") == 1 && options.count(p + "log_anti_quorum") == 1 &&
options.count(p + "storage_replicas") == 1 && options.count(p + "log_engine") == 1 &&
options.count(p + "storage_engine") == 1;
}
ACTOR Future<DatabaseConfiguration> getDatabaseConfiguration(Database cx) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
RangeResult res = wait(tr.getRange(configKeys, CLIENT_KNOBS->TOO_MANY));
ASSERT(res.size() < CLIENT_KNOBS->TOO_MANY);
DatabaseConfiguration config;
config.fromKeyValues((VectorRef<KeyValueRef>)res);
return config;
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<ConfigurationResult> changeConfig(Database cx, std::map<std::string, std::string> m, bool force) {
state StringRef initIdKey = LiteralStringRef("\xff/init_id");
state Transaction tr(cx);
if (!m.size()) {
return ConfigurationResult::NO_OPTIONS_PROVIDED;
}
// make sure we have essential configuration options
std::string initKey = configKeysPrefix.toString() + "initialized";
state bool creating = m.count(initKey) != 0;
state Optional<UID> locked;
{
auto iter = m.find(databaseLockedKey.toString());
if (iter != m.end()) {
if (!creating) {
return ConfigurationResult::LOCKED_NOT_NEW;
}
locked = UID::fromString(iter->second);
m.erase(iter);
}
}
if (creating) {
m[initIdKey.toString()] = deterministicRandom()->randomUniqueID().toString();
if (!isCompleteConfiguration(m)) {
return ConfigurationResult::INCOMPLETE_CONFIGURATION;
}
}
state Future<Void> tooLong = delay(60);
state Key versionKey = BinaryWriter::toValue(deterministicRandom()->randomUniqueID(), Unversioned());
state bool oldReplicationUsesDcId = false;
loop {
try {
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
if (!creating && !force) {
state Future<RangeResult> fConfig = tr.getRange(configKeys, CLIENT_KNOBS->TOO_MANY);
state Future<vector<ProcessData>> fWorkers = getWorkers(&tr);
wait(success(fConfig) || tooLong);
if (!fConfig.isReady()) {
return ConfigurationResult::DATABASE_UNAVAILABLE;
}
if (fConfig.isReady()) {
ASSERT(fConfig.get().size() < CLIENT_KNOBS->TOO_MANY);
state DatabaseConfiguration oldConfig;
oldConfig.fromKeyValues((VectorRef<KeyValueRef>)fConfig.get());
state DatabaseConfiguration newConfig = oldConfig;
for (auto kv : m) {
newConfig.set(kv.first, kv.second);
}
if (!newConfig.isValid()) {
return ConfigurationResult::INVALID_CONFIGURATION;
}
if (newConfig.tLogPolicy->attributeKeys().count("dcid") && newConfig.regions.size() > 0) {
return ConfigurationResult::REGION_REPLICATION_MISMATCH;
}
oldReplicationUsesDcId =
oldReplicationUsesDcId || oldConfig.tLogPolicy->attributeKeys().count("dcid");
if (oldConfig.usableRegions != newConfig.usableRegions) {
// cannot change region configuration
std::map<Key, int32_t> dcId_priority;
for (auto& it : newConfig.regions) {
dcId_priority[it.dcId] = it.priority;
}
for (auto& it : oldConfig.regions) {
if (!dcId_priority.count(it.dcId) || dcId_priority[it.dcId] != it.priority) {
return ConfigurationResult::REGIONS_CHANGED;
}
}
// must only have one region with priority >= 0
int activeRegionCount = 0;
for (auto& it : newConfig.regions) {
if (it.priority >= 0) {
activeRegionCount++;
}
}
if (activeRegionCount > 1) {
return ConfigurationResult::MULTIPLE_ACTIVE_REGIONS;
}
}
state Future<RangeResult> fServerList = (newConfig.regions.size())
? tr.getRange(serverListKeys, CLIENT_KNOBS->TOO_MANY)
: Future<RangeResult>();
if (newConfig.usableRegions == 2) {
if (oldReplicationUsesDcId) {
state Future<RangeResult> fLocalityList =
tr.getRange(tagLocalityListKeys, CLIENT_KNOBS->TOO_MANY);
wait(success(fLocalityList) || tooLong);
if (!fLocalityList.isReady()) {
return ConfigurationResult::DATABASE_UNAVAILABLE;
}
RangeResult localityList = fLocalityList.get();
ASSERT(!localityList.more && localityList.size() < CLIENT_KNOBS->TOO_MANY);
std::set<Key> localityDcIds;
for (auto& s : localityList) {
auto dc = decodeTagLocalityListKey(s.key);
if (dc.present()) {
localityDcIds.insert(dc.get());
}
}
for (auto& it : newConfig.regions) {
if (localityDcIds.count(it.dcId) == 0) {
return ConfigurationResult::DCID_MISSING;
}
}
} else {
// all regions with priority >= 0 must be fully replicated
state std::vector<Future<Optional<Value>>> replicasFutures;
for (auto& it : newConfig.regions) {
if (it.priority >= 0) {
replicasFutures.push_back(tr.get(datacenterReplicasKeyFor(it.dcId)));
}
}
wait(waitForAll(replicasFutures) || tooLong);
for (auto& it : replicasFutures) {
if (!it.isReady()) {
return ConfigurationResult::DATABASE_UNAVAILABLE;
}
if (!it.get().present()) {
return ConfigurationResult::REGION_NOT_FULLY_REPLICATED;
}
}
}
}
if (newConfig.regions.size()) {
// all storage servers must be in one of the regions
wait(success(fServerList) || tooLong);
if (!fServerList.isReady()) {
return ConfigurationResult::DATABASE_UNAVAILABLE;
}
RangeResult serverList = fServerList.get();
ASSERT(!serverList.more && serverList.size() < CLIENT_KNOBS->TOO_MANY);
std::set<Key> newDcIds;
for (auto& it : newConfig.regions) {
newDcIds.insert(it.dcId);
}
std::set<Optional<Key>> missingDcIds;
for (auto& s : serverList) {
auto ssi = decodeServerListValue(s.value);
if (!ssi.locality.dcId().present() || !newDcIds.count(ssi.locality.dcId().get())) {
missingDcIds.insert(ssi.locality.dcId());
}
}
if (missingDcIds.size() > (oldReplicationUsesDcId ? 1 : 0)) {
return ConfigurationResult::STORAGE_IN_UNKNOWN_DCID;
}
}
wait(success(fWorkers) || tooLong);
if (!fWorkers.isReady()) {
return ConfigurationResult::DATABASE_UNAVAILABLE;
}
if (newConfig.regions.size()) {
std::map<Optional<Key>, std::set<Optional<Key>>> dcId_zoneIds;
for (auto& it : fWorkers.get()) {
if (it.processClass.machineClassFitness(ProcessClass::Storage) <= ProcessClass::WorstFit) {
dcId_zoneIds[it.locality.dcId()].insert(it.locality.zoneId());
}
}
for (auto& region : newConfig.regions) {
if (dcId_zoneIds[region.dcId].size() <
std::max(newConfig.storageTeamSize, newConfig.tLogReplicationFactor)) {
return ConfigurationResult::NOT_ENOUGH_WORKERS;
}
if (region.satelliteTLogReplicationFactor > 0 && region.priority >= 0) {
int totalSatelliteProcesses = 0;
for (auto& sat : region.satellites) {
totalSatelliteProcesses += dcId_zoneIds[sat.dcId].size();
}
if (totalSatelliteProcesses < region.satelliteTLogReplicationFactor) {
return ConfigurationResult::NOT_ENOUGH_WORKERS;
}
}
}
} else {
std::set<Optional<Key>> zoneIds;
for (auto& it : fWorkers.get()) {
if (it.processClass.machineClassFitness(ProcessClass::Storage) <= ProcessClass::WorstFit) {
zoneIds.insert(it.locality.zoneId());
}
}
if (zoneIds.size() < std::max(newConfig.storageTeamSize, newConfig.tLogReplicationFactor)) {
return ConfigurationResult::NOT_ENOUGH_WORKERS;
}
}
}
}
if (creating) {
tr.setOption(FDBTransactionOptions::INITIALIZE_NEW_DATABASE);
tr.addReadConflictRange(singleKeyRange(initIdKey));
} else if (m.size()) {
// might be used in an emergency transaction, so make sure it is retry-self-conflicting and
// CAUSAL_WRITE_RISKY
tr.setOption(FDBTransactionOptions::CAUSAL_WRITE_RISKY);
tr.addReadConflictRange(singleKeyRange(m.begin()->first));
}
if (locked.present()) {
ASSERT(creating);
tr.atomicOp(databaseLockedKey,
BinaryWriter::toValue(locked.get(), Unversioned())
.withPrefix(LiteralStringRef("0123456789"))
.withSuffix(LiteralStringRef("\x00\x00\x00\x00")),
MutationRef::SetVersionstampedValue);
}
for (auto i = m.begin(); i != m.end(); ++i) {
tr.set(StringRef(i->first), StringRef(i->second));
}
tr.addReadConflictRange(singleKeyRange(moveKeysLockOwnerKey));
tr.set(moveKeysLockOwnerKey, versionKey);
wait(tr.commit());
break;
} catch (Error& e) {
state Error e1(e);
if ((e.code() == error_code_not_committed || e.code() == error_code_transaction_too_old) && creating) {
// The database now exists. Determine whether we created it or it was already existing/created by
// someone else. The latter is an error.
tr.reset();
loop {
try {
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
Optional<Value> v = wait(tr.get(initIdKey));
if (v != m[initIdKey.toString()])
return ConfigurationResult::DATABASE_ALREADY_CREATED;
else
return ConfigurationResult::DATABASE_CREATED;
} catch (Error& e2) {
wait(tr.onError(e2));
}
}
}
wait(tr.onError(e1));
}
}
return ConfigurationResult::SUCCESS;
}
ConfigureAutoResult parseConfig(StatusObject const& status) {
ConfigureAutoResult result;
StatusObjectReader statusObj(status);
StatusObjectReader statusObjCluster;
if (!statusObj.get("cluster", statusObjCluster))
return ConfigureAutoResult();
StatusObjectReader statusObjConfig;
if (!statusObjCluster.get("configuration", statusObjConfig))
return ConfigureAutoResult();
if (!statusObjConfig.get("redundancy.factor", result.old_replication))
return ConfigureAutoResult();
result.auto_replication = result.old_replication;
int storage_replication;
int log_replication;
if (result.old_replication == "single") {
result.auto_replication = "double";
storage_replication = 2;
log_replication = 2;
} else if (result.old_replication == "double" || result.old_replication == "fast_recovery_double") {
storage_replication = 2;
log_replication = 2;
} else if (result.old_replication == "triple" || result.old_replication == "fast_recovery_triple") {
storage_replication = 3;
log_replication = 3;
} else if (result.old_replication == "three_datacenter") {
storage_replication = 6;
log_replication = 4;
} else if (result.old_replication == "three_datacenter_fallback") {
storage_replication = 4;
log_replication = 4;
} else if (result.old_replication == "three_data_hall") {
storage_replication = 3;
log_replication = 4;
} else if (result.old_replication == "three_data_hall_fallback") {
storage_replication = 2;
log_replication = 4;
} else
return ConfigureAutoResult();
StatusObjectReader machinesMap;
if (!statusObjCluster.get("machines", machinesMap))
return ConfigureAutoResult();
std::map<std::string, std::string> machineid_dcid;
std::set<std::string> datacenters;
int machineCount = 0;
for (auto mach : machinesMap.obj()) {
StatusObjectReader machine(mach.second);
std::string dcId;
if (machine.get("datacenter_id", dcId)) {
machineid_dcid[mach.first] = dcId;
datacenters.insert(dcId);
}
machineCount++;
}
result.machines = machineCount;
if (datacenters.size() > 1)
return ConfigureAutoResult();
StatusObjectReader processesMap;
if (!statusObjCluster.get("processes", processesMap))
return ConfigureAutoResult();
std::set<std::string> oldMachinesWithTransaction;
int oldTransactionProcesses = 0;
std::map<std::string, std::vector<std::pair<NetworkAddress, ProcessClass>>> machine_processes;
int processCount = 0;
for (auto proc : processesMap.obj()) {
StatusObjectReader process(proc.second);
if (!process.has("excluded") || !process.last().get_bool()) {
std::string addrStr;
if (!process.get("address", addrStr))
return ConfigureAutoResult();
std::string class_source;
if (!process.get("class_source", class_source))
return ConfigureAutoResult();
std::string class_type;
if (!process.get("class_type", class_type))
return ConfigureAutoResult();
std::string machineId;
if (!process.get("machine_id", machineId))
return ConfigureAutoResult();
NetworkAddress addr = NetworkAddress::parse(addrStr);
ProcessClass processClass(class_type, class_source);
if (processClass.classType() == ProcessClass::TransactionClass ||
processClass.classType() == ProcessClass::LogClass) {
oldMachinesWithTransaction.insert(machineId);
}
if (processClass.classType() == ProcessClass::TransactionClass ||
processClass.classType() == ProcessClass::CommitProxyClass ||
processClass.classType() == ProcessClass::GrvProxyClass ||
processClass.classType() == ProcessClass::ResolutionClass ||
processClass.classType() == ProcessClass::StatelessClass ||
processClass.classType() == ProcessClass::LogClass) {
oldTransactionProcesses++;
}
if (processClass.classSource() == ProcessClass::AutoSource) {
processClass = ProcessClass(ProcessClass::UnsetClass, ProcessClass::CommandLineSource);
result.address_class[addr] = processClass;
}
if (processClass.classType() != ProcessClass::TesterClass) {
machine_processes[machineId].emplace_back(addr, processClass);
processCount++;
}
}
}
result.processes = processCount;
result.old_processes_with_transaction = oldTransactionProcesses;
result.old_machines_with_transaction = oldMachinesWithTransaction.size();
std::map<std::pair<int, std::string>, std::vector<std::pair<NetworkAddress, ProcessClass>>> count_processes;
for (auto& it : machine_processes) {
count_processes[std::make_pair(it.second.size(), it.first)] = it.second;
}
std::set<std::string> machinesWithTransaction;
std::set<std::string> machinesWithStorage;
int totalTransactionProcesses = 0;
int existingProxyCount = 0;
int existingGrvProxyCount = 0;
int existingResolverCount = 0;
int existingStatelessCount = 0;
for (auto& it : machine_processes) {
for (auto& proc : it.second) {
if (proc.second == ProcessClass::TransactionClass || proc.second == ProcessClass::LogClass) {
totalTransactionProcesses++;
machinesWithTransaction.insert(it.first);
}
if (proc.second == ProcessClass::StatelessClass) {
existingStatelessCount++;
}
if (proc.second == ProcessClass::CommitProxyClass) {
existingProxyCount++;
}
if (proc.second == ProcessClass::GrvProxyClass) {
existingGrvProxyCount++;
}
if (proc.second == ProcessClass::ResolutionClass) {
existingResolverCount++;
}
if (proc.second == ProcessClass::StorageClass) {
machinesWithStorage.insert(it.first);
}
if (proc.second == ProcessClass::UnsetClass && proc.second.classSource() == ProcessClass::DBSource) {
machinesWithStorage.insert(it.first);
}
}
}
if (processCount < 10)
return ConfigureAutoResult();
result.desired_resolvers = 1;
int resolverCount;
if (!statusObjConfig.get("resolvers", result.old_resolvers)) {
result.old_resolvers = CLIENT_KNOBS->DEFAULT_AUTO_RESOLVERS;
statusObjConfig.get("auto_resolvers", result.old_resolvers);
result.auto_resolvers = result.desired_resolvers;
resolverCount = result.auto_resolvers;
} else {
result.auto_resolvers = result.old_resolvers;
resolverCount = result.old_resolvers;
}
result.desired_commit_proxies = std::max(std::min(12, processCount / 15), 1);
int proxyCount;
if (!statusObjConfig.get("commit_proxies", result.old_commit_proxies)) {
result.old_commit_proxies = CLIENT_KNOBS->DEFAULT_AUTO_COMMIT_PROXIES;
statusObjConfig.get("auto_commit_proxies", result.old_commit_proxies);
result.auto_commit_proxies = result.desired_commit_proxies;
proxyCount = result.auto_commit_proxies;
} else {
result.auto_commit_proxies = result.old_commit_proxies;
proxyCount = result.old_commit_proxies;
}
result.desired_grv_proxies = std::max(std::min(4, processCount / 20), 1);
int grvProxyCount;
if (!statusObjConfig.get("grv_proxies", result.old_grv_proxies)) {
result.old_grv_proxies = CLIENT_KNOBS->DEFAULT_AUTO_GRV_PROXIES;
statusObjConfig.get("auto_grv_proxies", result.old_grv_proxies);
result.auto_grv_proxies = result.desired_grv_proxies;
grvProxyCount = result.auto_grv_proxies;
} else {
result.auto_grv_proxies = result.old_grv_proxies;
grvProxyCount = result.old_grv_proxies;
}
result.desired_logs = std::min(12, processCount / 20);
result.desired_logs = std::max(result.desired_logs, log_replication + 1);
result.desired_logs = std::min<int>(result.desired_logs, machine_processes.size());
int logCount;
if (!statusObjConfig.get("logs", result.old_logs)) {
result.old_logs = CLIENT_KNOBS->DEFAULT_AUTO_LOGS;
statusObjConfig.get("auto_logs", result.old_logs);
result.auto_logs = result.desired_logs;
logCount = result.auto_logs;
} else {
result.auto_logs = result.old_logs;
logCount = result.old_logs;
}
logCount = std::max(logCount, log_replication);
totalTransactionProcesses += std::min(existingProxyCount, proxyCount);
totalTransactionProcesses += std::min(existingGrvProxyCount, grvProxyCount);
totalTransactionProcesses += std::min(existingResolverCount, resolverCount);
totalTransactionProcesses += existingStatelessCount;
// if one process on a machine is transaction class, make them all transaction class
for (auto& it : count_processes) {
if (machinesWithTransaction.count(it.first.second) && !machinesWithStorage.count(it.first.second)) {
for (auto& proc : it.second) {
if (proc.second == ProcessClass::UnsetClass &&
proc.second.classSource() == ProcessClass::CommandLineSource) {
result.address_class[proc.first] =
ProcessClass(ProcessClass::TransactionClass, ProcessClass::AutoSource);
totalTransactionProcesses++;
}
}
}
}
int desiredTotalTransactionProcesses = logCount + resolverCount + proxyCount + grvProxyCount;
// add machines with all transaction class until we have enough processes and enough machines
for (auto& it : count_processes) {
if (machinesWithTransaction.size() >= logCount && totalTransactionProcesses >= desiredTotalTransactionProcesses)
break;
if (!machinesWithTransaction.count(it.first.second) && !machinesWithStorage.count(it.first.second)) {
for (auto& proc : it.second) {
if (proc.second == ProcessClass::UnsetClass &&
proc.second.classSource() == ProcessClass::CommandLineSource) {
ASSERT(proc.second != ProcessClass::TransactionClass);
result.address_class[proc.first] =
ProcessClass(ProcessClass::TransactionClass, ProcessClass::AutoSource);
totalTransactionProcesses++;
machinesWithTransaction.insert(it.first.second);
}
}
}
}
if (machinesWithTransaction.size() < logCount || totalTransactionProcesses < desiredTotalTransactionProcesses)
return ConfigureAutoResult();
result.auto_processes_with_transaction = totalTransactionProcesses;
result.auto_machines_with_transaction = machinesWithTransaction.size();
if (3 * totalTransactionProcesses > processCount)
return ConfigureAutoResult();
return result;
}
ACTOR Future<ConfigurationResult> autoConfig(Database cx, ConfigureAutoResult conf) {
state Transaction tr(cx);
state Key versionKey = BinaryWriter::toValue(deterministicRandom()->randomUniqueID(), Unversioned());
if (!conf.address_class.size())
return ConfigurationResult::INCOMPLETE_CONFIGURATION; // FIXME: correct return type
loop {
try {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
vector<ProcessData> workers = wait(getWorkers(&tr));
std::map<NetworkAddress, Optional<Standalone<StringRef>>> address_processId;
for (auto& w : workers) {
address_processId[w.address] = w.locality.processId();
}
for (auto& it : conf.address_class) {
if (it.second.classSource() == ProcessClass::CommandLineSource) {
tr.clear(processClassKeyFor(address_processId[it.first].get()));
} else {
tr.set(processClassKeyFor(address_processId[it.first].get()), processClassValue(it.second));
}
}
if (conf.address_class.size())
tr.set(processClassChangeKey, deterministicRandom()->randomUniqueID().toString());
if (conf.auto_logs != conf.old_logs)
tr.set(configKeysPrefix.toString() + "auto_logs", format("%d", conf.auto_logs));
if (conf.auto_commit_proxies != conf.old_commit_proxies)
tr.set(configKeysPrefix.toString() + "auto_commit_proxies", format("%d", conf.auto_commit_proxies));
if (conf.auto_grv_proxies != conf.old_grv_proxies)
tr.set(configKeysPrefix.toString() + "auto_grv_proxies", format("%d", conf.auto_grv_proxies));
if (conf.auto_resolvers != conf.old_resolvers)
tr.set(configKeysPrefix.toString() + "auto_resolvers", format("%d", conf.auto_resolvers));
if (conf.auto_replication != conf.old_replication) {
std::vector<StringRef> modes;
modes.push_back(conf.auto_replication);
std::map<std::string, std::string> m;
auto r = buildConfiguration(modes, m);
if (r != ConfigurationResult::SUCCESS)
return r;
for (auto& kv : m)
tr.set(kv.first, kv.second);
}
tr.addReadConflictRange(singleKeyRange(moveKeysLockOwnerKey));
tr.set(moveKeysLockOwnerKey, versionKey);
wait(tr.commit());
return ConfigurationResult::SUCCESS;
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
Future<ConfigurationResult> changeConfig(Database const& cx,
std::vector<StringRef> const& modes,
Optional<ConfigureAutoResult> const& conf,
bool force) {
if (modes.size() && modes[0] == LiteralStringRef("auto") && conf.present()) {
return autoConfig(cx, conf.get());
}
std::map<std::string, std::string> m;
auto r = buildConfiguration(modes, m);
if (r != ConfigurationResult::SUCCESS)
return r;
return changeConfig(cx, m, force);
}
Future<ConfigurationResult> changeConfig(Database const& cx, std::string const& modes, bool force) {
TraceEvent("ChangeConfig").detail("Mode", modes);
std::map<std::string, std::string> m;
auto r = buildConfiguration(modes, m);
if (r != ConfigurationResult::SUCCESS)
return r;
return changeConfig(cx, m, force);
}
ACTOR Future<vector<ProcessData>> getWorkers(Transaction* tr) {
state Future<RangeResult> processClasses = tr->getRange(processClassKeys, CLIENT_KNOBS->TOO_MANY);
state Future<RangeResult> processData = tr->getRange(workerListKeys, CLIENT_KNOBS->TOO_MANY);
wait(success(processClasses) && success(processData));
ASSERT(!processClasses.get().more && processClasses.get().size() < CLIENT_KNOBS->TOO_MANY);
ASSERT(!processData.get().more && processData.get().size() < CLIENT_KNOBS->TOO_MANY);
std::map<Optional<Standalone<StringRef>>, ProcessClass> id_class;
for (int i = 0; i < processClasses.get().size(); i++) {
id_class[decodeProcessClassKey(processClasses.get()[i].key)] =
decodeProcessClassValue(processClasses.get()[i].value);
}
std::vector<ProcessData> results;
for (int i = 0; i < processData.get().size(); i++) {
ProcessData data = decodeWorkerListValue(processData.get()[i].value);
ProcessClass processClass = id_class[data.locality.processId()];
if (processClass.classSource() == ProcessClass::DBSource ||
data.processClass.classType() == ProcessClass::UnsetClass)
data.processClass = processClass;
if (data.processClass.classType() != ProcessClass::TesterClass)
results.push_back(data);
}
return results;
}
ACTOR Future<vector<ProcessData>> getWorkers(Database cx) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::READ_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE); // necessary?
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
vector<ProcessData> workers = wait(getWorkers(&tr));
return workers;
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<std::vector<NetworkAddress>> getCoordinators(Database cx) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> currentKey = wait(tr.get(coordinatorsKey));
if (!currentKey.present())
return std::vector<NetworkAddress>();
return ClusterConnectionString(currentKey.get().toString()).coordinators();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Optional<CoordinatorsResult>> changeQuorumChecker(Transaction* tr,
Reference<IQuorumChange> change,
std::vector<NetworkAddress>* desiredCoordinators) {
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
tr->setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
tr->setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
Optional<Value> currentKey = wait(tr->get(coordinatorsKey));
if (!currentKey.present())
return CoordinatorsResult::BAD_DATABASE_STATE; // Someone deleted this key entirely?
state ClusterConnectionString old(currentKey.get().toString());
if (tr->getDatabase()->getConnectionFile() &&
old.clusterKeyName().toString() !=
tr->getDatabase()->getConnectionFile()->getConnectionString().clusterKeyName())
return CoordinatorsResult::BAD_DATABASE_STATE; // Someone changed the "name" of the database??
state CoordinatorsResult result = CoordinatorsResult::SUCCESS;
if (!desiredCoordinators->size()) {
std::vector<NetworkAddress> _desiredCoordinators = wait(change->getDesiredCoordinators(
tr, old.coordinators(), Reference<ClusterConnectionFile>(new ClusterConnectionFile(old)), result));
*desiredCoordinators = _desiredCoordinators;
}
if (result != CoordinatorsResult::SUCCESS)
return result;
if (!desiredCoordinators->size())
return CoordinatorsResult::INVALID_NETWORK_ADDRESSES;
std::sort(desiredCoordinators->begin(), desiredCoordinators->end());
std::string newName = change->getDesiredClusterKeyName();
if (newName.empty())
newName = old.clusterKeyName().toString();
if (old.coordinators() == *desiredCoordinators && old.clusterKeyName() == newName)
return CoordinatorsResult::SAME_NETWORK_ADDRESSES;
state ClusterConnectionString conn(*desiredCoordinators,
StringRef(newName + ':' + deterministicRandom()->randomAlphaNumeric(32)));
if (g_network->isSimulated()) {
for (int i = 0; i < (desiredCoordinators->size() / 2) + 1; i++) {
auto addresses = g_simulator.getProcessByAddress((*desiredCoordinators)[i])->addresses;
g_simulator.protectedAddresses.insert(addresses.address);
if (addresses.secondaryAddress.present()) {
g_simulator.protectedAddresses.insert(addresses.secondaryAddress.get());
}
TraceEvent("ProtectCoordinator").detail("Address", (*desiredCoordinators)[i]).backtrace();
}
}
vector<Future<Optional<LeaderInfo>>> leaderServers;
ClientCoordinators coord(Reference<ClusterConnectionFile>(new ClusterConnectionFile(conn)));
leaderServers.reserve(coord.clientLeaderServers.size());
for (int i = 0; i < coord.clientLeaderServers.size(); i++)
leaderServers.push_back(retryBrokenPromise(coord.clientLeaderServers[i].getLeader,
GetLeaderRequest(coord.clusterKey, UID()),
TaskPriority::CoordinationReply));
choose {
when(wait(waitForAll(leaderServers))) {}
when(wait(delay(5.0))) { return CoordinatorsResult::COORDINATOR_UNREACHABLE; }
}
tr->set(coordinatorsKey, conn.toString());
return Optional<CoordinatorsResult>();
}
ACTOR Future<CoordinatorsResult> changeQuorum(Database cx, Reference<IQuorumChange> change) {
state Transaction tr(cx);
state int retries = 0;
state std::vector<NetworkAddress> desiredCoordinators;
state int notEnoughMachineResults = 0;
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
Optional<Value> currentKey = wait(tr.get(coordinatorsKey));
if (!currentKey.present())
return CoordinatorsResult::BAD_DATABASE_STATE; // Someone deleted this key entirely?
state ClusterConnectionString old(currentKey.get().toString());
if (cx->getConnectionFile() &&
old.clusterKeyName().toString() != cx->getConnectionFile()->getConnectionString().clusterKeyName())
return CoordinatorsResult::BAD_DATABASE_STATE; // Someone changed the "name" of the database??
state CoordinatorsResult result = CoordinatorsResult::SUCCESS;
if (!desiredCoordinators.size()) {
std::vector<NetworkAddress> _desiredCoordinators = wait(change->getDesiredCoordinators(
&tr, old.coordinators(), Reference<ClusterConnectionFile>(new ClusterConnectionFile(old)), result));
desiredCoordinators = _desiredCoordinators;
}
if (result == CoordinatorsResult::NOT_ENOUGH_MACHINES && notEnoughMachineResults < 1) {
// we could get not_enough_machines if we happen to see the database while the cluster controller is
// updating the worker list, so make sure it happens twice before returning a failure
notEnoughMachineResults++;
wait(delay(1.0));
tr.reset();
continue;
}
if (result != CoordinatorsResult::SUCCESS)
return result;
if (!desiredCoordinators.size())
return CoordinatorsResult::INVALID_NETWORK_ADDRESSES;
std::sort(desiredCoordinators.begin(), desiredCoordinators.end());
std::string newName = change->getDesiredClusterKeyName();
if (newName.empty())
newName = old.clusterKeyName().toString();
if (old.coordinators() == desiredCoordinators && old.clusterKeyName() == newName)
return retries ? CoordinatorsResult::SUCCESS : CoordinatorsResult::SAME_NETWORK_ADDRESSES;
state ClusterConnectionString conn(
desiredCoordinators, StringRef(newName + ':' + deterministicRandom()->randomAlphaNumeric(32)));
if (g_network->isSimulated()) {
for (int i = 0; i < (desiredCoordinators.size() / 2) + 1; i++) {
auto process = g_simulator.getProcessByAddress(desiredCoordinators[i]);
ASSERT(process->isReliable() || process->rebooting);
g_simulator.protectedAddresses.insert(process->addresses.address);
if (process->addresses.secondaryAddress.present()) {
g_simulator.protectedAddresses.insert(process->addresses.secondaryAddress.get());
}
TraceEvent("ProtectCoordinator").detail("Address", desiredCoordinators[i]).backtrace();
}
}
TraceEvent("AttemptingQuorumChange").detail("FromCS", old.toString()).detail("ToCS", conn.toString());
TEST(old.clusterKeyName() != conn.clusterKeyName()); // Quorum change with new name
TEST(old.clusterKeyName() == conn.clusterKeyName()); // Quorum change with unchanged name
state vector<Future<Optional<LeaderInfo>>> leaderServers;
state ClientCoordinators coord(Reference<ClusterConnectionFile>(new ClusterConnectionFile(conn)));
// check if allowed to modify the cluster descriptor
if (!change->getDesiredClusterKeyName().empty()) {
CheckDescriptorMutableReply mutabilityReply =
wait(coord.clientLeaderServers[0].checkDescriptorMutable.getReply(CheckDescriptorMutableRequest()));
if (!mutabilityReply.isMutable)
return CoordinatorsResult::BAD_DATABASE_STATE;
}
leaderServers.reserve(coord.clientLeaderServers.size());
for (int i = 0; i < coord.clientLeaderServers.size(); i++)
leaderServers.push_back(retryBrokenPromise(coord.clientLeaderServers[i].getLeader,
GetLeaderRequest(coord.clusterKey, UID()),
TaskPriority::CoordinationReply));
choose {
when(wait(waitForAll(leaderServers))) {}
when(wait(delay(5.0))) { return CoordinatorsResult::COORDINATOR_UNREACHABLE; }
}
tr.set(coordinatorsKey, conn.toString());
wait(tr.commit());
ASSERT(false); // commit should fail, but the value has changed
} catch (Error& e) {
TraceEvent("RetryQuorumChange").error(e).detail("Retries", retries);
wait(tr.onError(e));
++retries;
}
}
}
struct SpecifiedQuorumChange final : IQuorumChange {
vector<NetworkAddress> desired;
explicit SpecifiedQuorumChange(vector<NetworkAddress> const& desired) : desired(desired) {}
Future<vector<NetworkAddress>> getDesiredCoordinators(Transaction* tr,
vector<NetworkAddress> oldCoordinators,
Reference<ClusterConnectionFile>,
CoordinatorsResult&) override {
return desired;
}
};
Reference<IQuorumChange> specifiedQuorumChange(vector<NetworkAddress> const& addresses) {
return Reference<IQuorumChange>(new SpecifiedQuorumChange(addresses));
}
struct NoQuorumChange final : IQuorumChange {
Future<vector<NetworkAddress>> getDesiredCoordinators(Transaction* tr,
vector<NetworkAddress> oldCoordinators,
Reference<ClusterConnectionFile>,
CoordinatorsResult&) override {
return oldCoordinators;
}
};
Reference<IQuorumChange> noQuorumChange() {
return Reference<IQuorumChange>(new NoQuorumChange);
}
struct NameQuorumChange final : IQuorumChange {
std::string newName;
Reference<IQuorumChange> otherChange;
explicit NameQuorumChange(std::string const& newName, Reference<IQuorumChange> const& otherChange)
: newName(newName), otherChange(otherChange) {}
Future<vector<NetworkAddress>> getDesiredCoordinators(Transaction* tr,
vector<NetworkAddress> oldCoordinators,
Reference<ClusterConnectionFile> cf,
CoordinatorsResult& t) override {
return otherChange->getDesiredCoordinators(tr, oldCoordinators, cf, t);
}
std::string getDesiredClusterKeyName() const override { return newName; }
};
Reference<IQuorumChange> nameQuorumChange(std::string const& name, Reference<IQuorumChange> const& other) {
return Reference<IQuorumChange>(new NameQuorumChange(name, other));
}
struct AutoQuorumChange final : IQuorumChange {
int desired;
explicit AutoQuorumChange(int desired) : desired(desired) {}
Future<vector<NetworkAddress>> getDesiredCoordinators(Transaction* tr,
vector<NetworkAddress> oldCoordinators,
Reference<ClusterConnectionFile> ccf,
CoordinatorsResult& err) override {
return getDesired(Reference<AutoQuorumChange>::addRef(this), tr, oldCoordinators, ccf, &err);
}
ACTOR static Future<int> getRedundancy(AutoQuorumChange* self, Transaction* tr) {
state Future<Optional<Value>> fStorageReplicas =
tr->get(LiteralStringRef("storage_replicas").withPrefix(configKeysPrefix));
state Future<Optional<Value>> fLogReplicas =
tr->get(LiteralStringRef("log_replicas").withPrefix(configKeysPrefix));
wait(success(fStorageReplicas) && success(fLogReplicas));
int redundancy = std::min(atoi(fStorageReplicas.get().get().toString().c_str()),
atoi(fLogReplicas.get().get().toString().c_str()));
return redundancy;
}
ACTOR static Future<bool> isAcceptable(AutoQuorumChange* self,
Transaction* tr,
vector<NetworkAddress> oldCoordinators,
Reference<ClusterConnectionFile> ccf,
int desiredCount,
std::set<AddressExclusion>* excluded) {
// Are there enough coordinators for the redundancy level?
if (oldCoordinators.size() < desiredCount)
return false;
if (oldCoordinators.size() % 2 != 1)
return false;
// Check availability
ClientCoordinators coord(ccf);
vector<Future<Optional<LeaderInfo>>> leaderServers;
leaderServers.reserve(coord.clientLeaderServers.size());
for (int i = 0; i < coord.clientLeaderServers.size(); i++) {
leaderServers.push_back(retryBrokenPromise(coord.clientLeaderServers[i].getLeader,
GetLeaderRequest(coord.clusterKey, UID()),
TaskPriority::CoordinationReply));
}
Optional<vector<Optional<LeaderInfo>>> results =
wait(timeout(getAll(leaderServers), CLIENT_KNOBS->IS_ACCEPTABLE_DELAY));
if (!results.present()) {
return false;
} // Not all responded
for (auto& r : results.get()) {
if (!r.present()) {
return false; // Coordinator doesn't know about this database?
}
}
// Check exclusions
for (auto& c : oldCoordinators) {
if (addressExcluded(*excluded, c))
return false;
}
// Check locality
// FIXME: Actual locality!
std::sort(oldCoordinators.begin(), oldCoordinators.end());
for (int i = 1; i < oldCoordinators.size(); i++)
if (oldCoordinators[i - 1].ip == oldCoordinators[i].ip)
return false; // Multiple coordinators share an IP
return true; // The status quo seems fine
}
ACTOR static Future<vector<NetworkAddress>> getDesired(Reference<AutoQuorumChange> self,
Transaction* tr,
vector<NetworkAddress> oldCoordinators,
Reference<ClusterConnectionFile> ccf,
CoordinatorsResult* err) {
state int desiredCount = self->desired;
if (desiredCount == -1) {
int redundancy = wait(getRedundancy(self.getPtr(), tr));
desiredCount = redundancy * 2 - 1;
}
std::vector<AddressExclusion> excl = wait(getExcludedServers(tr));
state std::set<AddressExclusion> excluded(excl.begin(), excl.end());
vector<ProcessData> _workers = wait(getWorkers(tr));
state vector<ProcessData> workers = _workers;
std::map<NetworkAddress, LocalityData> addr_locality;
for (auto w : workers)
addr_locality[w.address] = w.locality;
// since we don't have the locality data for oldCoordinators:
// check if every old coordinator is in the workers vector and
// check if multiple old coordinators map to the same locality data (same machine)
bool checkAcceptable = true;
std::set<Optional<Standalone<StringRef>>> checkDuplicates;
for (auto addr : oldCoordinators) {
auto findResult = addr_locality.find(addr);
if (findResult == addr_locality.end() || checkDuplicates.count(findResult->second.zoneId())) {
checkAcceptable = false;
break;
}
checkDuplicates.insert(findResult->second.zoneId());
}
if (checkAcceptable) {
bool ok = wait(isAcceptable(self.getPtr(), tr, oldCoordinators, ccf, desiredCount, &excluded));
if (ok)
return oldCoordinators;
}
std::vector<NetworkAddress> chosen;
self->addDesiredWorkers(chosen, workers, desiredCount, excluded);
if (chosen.size() < desiredCount) {
if (chosen.size() < oldCoordinators.size()) {
TraceEvent("NotEnoughMachinesForCoordinators")
.detail("EligibleWorkers", workers.size())
.detail("DesiredCoordinators", desiredCount)
.detail("CurrentCoordinators", oldCoordinators.size());
*err = CoordinatorsResult::NOT_ENOUGH_MACHINES;
return vector<NetworkAddress>();
}
chosen.resize((chosen.size() - 1) | 1);
}
return chosen;
}
// Select a desired set of workers such that
// (1) the number of workers at each locality type (e.g., dcid) <= desiredCount; and
// (2) prefer workers at a locality where less workers has been chosen than other localities: evenly distribute
// workers.
void addDesiredWorkers(vector<NetworkAddress>& chosen,
const vector<ProcessData>& workers,
int desiredCount,
const std::set<AddressExclusion>& excluded) {
vector<ProcessData> remainingWorkers(workers);
deterministicRandom()->randomShuffle(remainingWorkers);
std::partition(remainingWorkers.begin(), remainingWorkers.end(), [](const ProcessData& data) {
return (data.processClass == ProcessClass::CoordinatorClass);
});
TraceEvent(SevDebug, "AutoSelectCoordinators").detail("CandidateWorkers", remainingWorkers.size());
for (auto worker = remainingWorkers.begin(); worker != remainingWorkers.end(); worker++) {
TraceEvent(SevDebug, "AutoSelectCoordinators")
.detail("Worker", worker->processClass.toString())
.detail("Address", worker->address.toString())
.detail("Locality", worker->locality.toString());
}
TraceEvent(SevDebug, "AutoSelectCoordinators").detail("ExcludedAddress", excluded.size());
for (auto& excludedAddr : excluded) {
TraceEvent(SevDebug, "AutoSelectCoordinators").detail("ExcludedAddress", excludedAddr.toString());
}
std::map<StringRef, int> maxCounts;
std::map<StringRef, std::map<StringRef, int>> currentCounts;
std::map<StringRef, int> hardLimits;
vector<StringRef> fields({ LiteralStringRef("dcid"),
LiteralStringRef("data_hall"),
LiteralStringRef("zoneid"),
LiteralStringRef("machineid") });
for (auto field = fields.begin(); field != fields.end(); field++) {
if (field->toString() == "zoneid") {
hardLimits[*field] = 1;
} else {
hardLimits[*field] = desiredCount;
}
}
while (chosen.size() < desiredCount) {
bool found = false;
for (auto worker = remainingWorkers.begin(); worker != remainingWorkers.end(); worker++) {
if (addressExcluded(excluded, worker->address)) {
continue;
}
// Exclude faulty node due to machine assassination
if (g_network->isSimulated() && !g_simulator.getProcessByAddress(worker->address)->isReliable()) {
TraceEvent("AutoSelectCoordinators").detail("SkipUnreliableWorker", worker->address.toString());
continue;
}
bool valid = true;
for (auto field = fields.begin(); field != fields.end(); field++) {
if (maxCounts[*field] == 0) {
maxCounts[*field] = 1;
}
auto value = worker->locality.get(*field).orDefault(LiteralStringRef(""));
auto currentCount = currentCounts[*field][value];
if (currentCount >= maxCounts[*field]) {
valid = false;
break;
}
}
if (valid) {
for (auto field = fields.begin(); field != fields.end(); field++) {
auto value = worker->locality.get(*field).orDefault(LiteralStringRef(""));
currentCounts[*field][value] += 1;
}
chosen.push_back(worker->address);
remainingWorkers.erase(worker);
found = true;
break;
}
}
if (!found) {
bool canIncrement = false;
for (auto field = fields.begin(); field != fields.end(); field++) {
if (maxCounts[*field] < hardLimits[*field]) {
maxCounts[*field] += 1;
canIncrement = true;
break;
}
}
if (!canIncrement) {
break;
}
}
}
}
};
Reference<IQuorumChange> autoQuorumChange(int desired) {
return Reference<IQuorumChange>(new AutoQuorumChange(desired));
}
void excludeServers(Transaction& tr, vector<AddressExclusion>& servers, bool failed) {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
std::string excludeVersionKey = deterministicRandom()->randomUniqueID().toString();
auto serversVersionKey = failed ? failedServersVersionKey : excludedServersVersionKey;
tr.addReadConflictRange(singleKeyRange(serversVersionKey)); // To conflict with parallel includeServers
tr.set(serversVersionKey, excludeVersionKey);
for (auto& s : servers) {
if (failed) {
tr.set(encodeFailedServersKey(s), StringRef());
} else {
tr.set(encodeExcludedServersKey(s), StringRef());
}
}
TraceEvent("ExcludeServersCommit").detail("Servers", describe(servers)).detail("ExcludeFailed", failed);
}
ACTOR Future<Void> excludeServers(Database cx, vector<AddressExclusion> servers, bool failed) {
if (cx->apiVersionAtLeast(700)) {
state ReadYourWritesTransaction ryw(cx);
loop {
try {
ryw.setOption(FDBTransactionOptions::SPECIAL_KEY_SPACE_ENABLE_WRITES);
ryw.set(
SpecialKeySpace::getManagementApiCommandOptionSpecialKey(failed ? "failed" : "excluded", "force"),
ValueRef());
for (auto& s : servers) {
Key addr = failed
? SpecialKeySpace::getManagementApiCommandPrefix("failed").withSuffix(s.toString())
: SpecialKeySpace::getManagementApiCommandPrefix("exclude").withSuffix(s.toString());
ryw.set(addr, ValueRef());
}
TraceEvent("ExcludeServersSpecialKeySpaceCommit")
.detail("Servers", describe(servers))
.detail("ExcludeFailed", failed);
wait(ryw.commit());
return Void();
} catch (Error& e) {
wait(ryw.onError(e));
}
}
} else {
state Transaction tr(cx);
loop {
try {
excludeServers(tr, servers, failed);
wait(tr.commit());
return Void();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
}
ACTOR Future<Void> includeServers(Database cx, vector<AddressExclusion> servers, bool failed) {
state std::string versionKey = deterministicRandom()->randomUniqueID().toString();
if (cx->apiVersionAtLeast(700)) {
state ReadYourWritesTransaction ryw(cx);
loop {
try {
ryw.setOption(FDBTransactionOptions::SPECIAL_KEY_SPACE_ENABLE_WRITES);
for (auto& s : servers) {
if (!s.isValid()) {
if (failed) {
ryw.clear(SpecialKeySpace::getManamentApiCommandRange("failed"));
} else {
ryw.clear(SpecialKeySpace::getManamentApiCommandRange("exclude"));
}
} else {
Key addr =
failed ? SpecialKeySpace::getManagementApiCommandPrefix("failed").withSuffix(s.toString())
: SpecialKeySpace::getManagementApiCommandPrefix("exclude").withSuffix(s.toString());
ryw.clear(addr);
// Eliminate both any ip-level exclusion (1.2.3.4) and any
// port-level exclusions (1.2.3.4:5)
// The range ['IP', 'IP;'] was originally deleted. ';' is
// char(':' + 1). This does not work, as other for all
// x between 0 and 9, 'IPx' will also be in this range.
//
// This is why we now make two clears: first only of the ip
// address, the second will delete all ports.
if (s.isWholeMachine())
ryw.clear(KeyRangeRef(addr.withSuffix(LiteralStringRef(":")),
addr.withSuffix(LiteralStringRef(";"))));
}
}
TraceEvent("IncludeServersCommit").detail("Servers", describe(servers)).detail("Failed", failed);
wait(ryw.commit());
return Void();
} catch (Error& e) {
TraceEvent("IncludeServersError").error(e, true);
wait(ryw.onError(e));
}
}
} else {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
// includeServers might be used in an emergency transaction, so make sure it is retry-self-conflicting
// and CAUSAL_WRITE_RISKY
tr.setOption(FDBTransactionOptions::CAUSAL_WRITE_RISKY);
if (failed) {
tr.addReadConflictRange(singleKeyRange(failedServersVersionKey));
tr.set(failedServersVersionKey, versionKey);
} else {
tr.addReadConflictRange(singleKeyRange(excludedServersVersionKey));
tr.set(excludedServersVersionKey, versionKey);
}
for (auto& s : servers) {
if (!s.isValid()) {
if (failed) {
tr.clear(failedServersKeys);
} else {
tr.clear(excludedServersKeys);
}
} else if (s.isWholeMachine()) {
// Eliminate both any ip-level exclusion (1.2.3.4) and any
// port-level exclusions (1.2.3.4:5)
// The range ['IP', 'IP;'] was originally deleted. ';' is
// char(':' + 1). This does not work, as other for all
// x between 0 and 9, 'IPx' will also be in this range.
//
// This is why we now make two clears: first only of the ip
// address, the second will delete all ports.
auto addr = failed ? encodeFailedServersKey(s) : encodeExcludedServersKey(s);
tr.clear(singleKeyRange(addr));
tr.clear(KeyRangeRef(addr + ':', addr + char(':' + 1)));
} else {
if (failed) {
tr.clear(encodeFailedServersKey(s));
} else {
tr.clear(encodeExcludedServersKey(s));
}
}
}
TraceEvent("IncludeServersCommit").detail("Servers", describe(servers)).detail("Failed", failed);
wait(tr.commit());
return Void();
} catch (Error& e) {
TraceEvent("IncludeServersError").error(e, true);
wait(tr.onError(e));
}
}
}
}
ACTOR Future<Void> setClass(Database cx, AddressExclusion server, ProcessClass processClass) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::USE_PROVISIONAL_PROXIES);
vector<ProcessData> workers = wait(getWorkers(&tr));
bool foundChange = false;
for (int i = 0; i < workers.size(); i++) {
if (server.excludes(workers[i].address)) {
if (processClass.classType() != ProcessClass::InvalidClass)
tr.set(processClassKeyFor(workers[i].locality.processId().get()),
processClassValue(processClass));
else
tr.clear(processClassKeyFor(workers[i].locality.processId().get()));
foundChange = true;
}
}
if (foundChange)
tr.set(processClassChangeKey, deterministicRandom()->randomUniqueID().toString());
wait(tr.commit());
return Void();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<vector<AddressExclusion>> getExcludedServers(Transaction* tr) {
state RangeResult r = wait(tr->getRange(excludedServersKeys, CLIENT_KNOBS->TOO_MANY));
ASSERT(!r.more && r.size() < CLIENT_KNOBS->TOO_MANY);
state RangeResult r2 = wait(tr->getRange(failedServersKeys, CLIENT_KNOBS->TOO_MANY));
ASSERT(!r2.more && r2.size() < CLIENT_KNOBS->TOO_MANY);
vector<AddressExclusion> exclusions;
for (auto i = r.begin(); i != r.end(); ++i) {
auto a = decodeExcludedServersKey(i->key);
if (a.isValid())
exclusions.push_back(a);
}
for (auto i = r2.begin(); i != r2.end(); ++i) {
auto a = decodeFailedServersKey(i->key);
if (a.isValid())
exclusions.push_back(a);
}
uniquify(exclusions);
return exclusions;
}
ACTOR Future<vector<AddressExclusion>> getExcludedServers(Database cx) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::READ_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE); // necessary?
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
vector<AddressExclusion> exclusions = wait(getExcludedServers(&tr));
return exclusions;
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> printHealthyZone(Database cx) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr.get(healthyZoneKey));
if (val.present() && decodeHealthyZoneValue(val.get()).first == ignoreSSFailuresZoneString) {
printf("Data distribution has been disabled for all storage server failures in this cluster and thus "
"maintenance mode is not active.\n");
} else if (!val.present() || decodeHealthyZoneValue(val.get()).second <= tr.getReadVersion().get()) {
printf("No ongoing maintenance.\n");
} else {
auto healthyZone = decodeHealthyZoneValue(val.get());
printf("Maintenance for zone %s will continue for %" PRId64 " seconds.\n",
healthyZone.first.toString().c_str(),
(healthyZone.second - tr.getReadVersion().get()) / CLIENT_KNOBS->CORE_VERSIONSPERSECOND);
}
return Void();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<bool> clearHealthyZone(Database cx, bool printWarning, bool clearSSFailureZoneString) {
state Transaction tr(cx);
TraceEvent("ClearHealthyZone").detail("ClearSSFailureZoneString", clearSSFailureZoneString);
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
Optional<Value> val = wait(tr.get(healthyZoneKey));
if (!clearSSFailureZoneString && val.present() &&
decodeHealthyZoneValue(val.get()).first == ignoreSSFailuresZoneString) {
if (printWarning) {
printf("ERROR: Maintenance mode cannot be used while data distribution is disabled for storage "
"server failures. Use 'datadistribution on' to reenable data distribution.\n");
}
return false;
}
tr.clear(healthyZoneKey);
wait(tr.commit());
return true;
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<bool> setHealthyZone(Database cx, StringRef zoneId, double seconds, bool printWarning) {
state Transaction tr(cx);
TraceEvent("SetHealthyZone").detail("Zone", zoneId).detail("DurationSeconds", seconds);
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
Optional<Value> val = wait(tr.get(healthyZoneKey));
if (val.present() && decodeHealthyZoneValue(val.get()).first == ignoreSSFailuresZoneString) {
if (printWarning) {
printf("ERROR: Maintenance mode cannot be used while data distribution is disabled for storage "
"server failures. Use 'datadistribution on' to reenable data distribution.\n");
}
return false;
}
Version readVersion = wait(tr.getReadVersion());
tr.set(healthyZoneKey,
healthyZoneValue(zoneId, readVersion + (seconds * CLIENT_KNOBS->CORE_VERSIONSPERSECOND)));
wait(tr.commit());
return true;
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> setDDIgnoreRebalanceSwitch(Database cx, bool ignoreRebalance) {
state Transaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
if (ignoreRebalance) {
tr.set(rebalanceDDIgnoreKey, LiteralStringRef("on"));
} else {
tr.clear(rebalanceDDIgnoreKey);
}
wait(tr.commit());
return Void();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<int> setDDMode(Database cx, int mode) {
state Transaction tr(cx);
state int oldMode = -1;
state BinaryWriter wr(Unversioned());
wr << mode;
loop {
try {
Optional<Value> old = wait(tr.get(dataDistributionModeKey));
if (oldMode < 0) {
oldMode = 1;
if (old.present()) {
BinaryReader rd(old.get(), Unversioned());
rd >> oldMode;
}
}
BinaryWriter wrMyOwner(Unversioned());
wrMyOwner << dataDistributionModeLock;
tr.set(moveKeysLockOwnerKey, wrMyOwner.toValue());
BinaryWriter wrLastWrite(Unversioned());
wrLastWrite << deterministicRandom()->randomUniqueID();
tr.set(moveKeysLockWriteKey, wrLastWrite.toValue());
tr.set(dataDistributionModeKey, wr.toValue());
if (mode) {
// set DDMode to 1 will enable all disabled parts, for instance the SS failure monitors.
Optional<Value> currentHealthyZoneValue = wait(tr.get(healthyZoneKey));
if (currentHealthyZoneValue.present() &&
decodeHealthyZoneValue(currentHealthyZoneValue.get()).first == ignoreSSFailuresZoneString) {
// only clear the key if it is currently being used to disable all SS failure data movement
tr.clear(healthyZoneKey);
}
tr.clear(rebalanceDDIgnoreKey);
}
wait(tr.commit());
return oldMode;
} catch (Error& e) {
TraceEvent("SetDDModeRetrying").error(e);
wait(tr.onError(e));
}
}
}
ACTOR Future<bool> checkForExcludingServersTxActor(ReadYourWritesTransaction* tr,
std::set<AddressExclusion>* exclusions,
std::set<NetworkAddress>* inProgressExclusion) {
// TODO : replace using ExclusionInProgressRangeImpl in special key space
ASSERT(inProgressExclusion->size() == 0); // Make sure every time it is cleared beforehand
if (!exclusions->size())
return true;
tr->setOption(FDBTransactionOptions::READ_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE); // necessary?
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
// Just getting a consistent read version proves that a set of tlogs satisfying the exclusions has completed
// recovery
// Check that there aren't any storage servers with addresses violating the exclusions
RangeResult serverList = wait(tr->getRange(serverListKeys, CLIENT_KNOBS->TOO_MANY));
ASSERT(!serverList.more && serverList.size() < CLIENT_KNOBS->TOO_MANY);
state bool ok = true;
for (auto& s : serverList) {
auto addresses = decodeServerListValue(s.value).getKeyValues.getEndpoint().addresses;
if (addressExcluded(*exclusions, addresses.address)) {
ok = false;
inProgressExclusion->insert(addresses.address);
}
if (addresses.secondaryAddress.present() && addressExcluded(*exclusions, addresses.secondaryAddress.get())) {
ok = false;
inProgressExclusion->insert(addresses.secondaryAddress.get());
}
}
if (ok) {
Optional<Standalone<StringRef>> value = wait(tr->get(logsKey));
ASSERT(value.present());
auto logs = decodeLogsValue(value.get());
for (auto const& log : logs.first) {
if (log.second == NetworkAddress() || addressExcluded(*exclusions, log.second)) {
ok = false;
inProgressExclusion->insert(log.second);
}
}
for (auto const& log : logs.second) {
if (log.second == NetworkAddress() || addressExcluded(*exclusions, log.second)) {
ok = false;
inProgressExclusion->insert(log.second);
}
}
}
return ok;
}
ACTOR Future<std::set<NetworkAddress>> checkForExcludingServers(Database cx,
vector<AddressExclusion> excl,
bool waitForAllExcluded) {
state std::set<AddressExclusion> exclusions(excl.begin(), excl.end());
state std::set<NetworkAddress> inProgressExclusion;
loop {
state ReadYourWritesTransaction tr(cx);
inProgressExclusion.clear();
try {
bool ok = wait(checkForExcludingServersTxActor(&tr, &exclusions, &inProgressExclusion));
if (ok)
return inProgressExclusion;
if (!waitForAllExcluded)
break;
wait(delayJittered(1.0)); // SOMEDAY: watches!
} catch (Error& e) {
TraceEvent("CheckForExcludingServersError").error(e);
wait(tr.onError(e));
}
}
return inProgressExclusion;
}
ACTOR Future<Void> mgmtSnapCreate(Database cx, Standalone<StringRef> snapCmd, UID snapUID) {
try {
wait(snapCreate(cx, snapCmd, snapUID));
TraceEvent("SnapCreateSucceeded").detail("snapUID", snapUID);
return Void();
} catch (Error& e) {
TraceEvent(SevWarn, "SnapCreateFailed").detail("snapUID", snapUID).error(e);
throw;
}
}
ACTOR Future<Void> waitForFullReplication(Database cx) {
state ReadYourWritesTransaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::READ_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::PRIORITY_SYSTEM_IMMEDIATE);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
RangeResult confResults = wait(tr.getRange(configKeys, CLIENT_KNOBS->TOO_MANY));
ASSERT(!confResults.more && confResults.size() < CLIENT_KNOBS->TOO_MANY);
state DatabaseConfiguration config;
config.fromKeyValues((VectorRef<KeyValueRef>)confResults);
state std::vector<Future<Optional<Value>>> replicasFutures;
for (auto& region : config.regions) {
replicasFutures.push_back(tr.get(datacenterReplicasKeyFor(region.dcId)));
}
wait(waitForAll(replicasFutures));
state std::vector<Future<Void>> watchFutures;
for (int i = 0; i < config.regions.size(); i++) {
if (!replicasFutures[i].get().present() ||
decodeDatacenterReplicasValue(replicasFutures[i].get().get()) < config.storageTeamSize) {
watchFutures.push_back(tr.watch(datacenterReplicasKeyFor(config.regions[i].dcId)));
}
}
if (!watchFutures.size() || (config.usableRegions == 1 && watchFutures.size() < config.regions.size())) {
return Void();
}
wait(tr.commit());
wait(waitForAny(watchFutures));
tr.reset();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> timeKeeperSetDisable(Database cx) {
loop {
state Transaction tr(cx);
try {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.set(timeKeeperDisableKey, StringRef());
wait(tr.commit());
return Void();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> lockDatabase(Transaction* tr, UID id) {
tr->setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr->get(databaseLockedKey));
if (val.present()) {
if (BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()) == id) {
return Void();
} else {
//TraceEvent("DBA_LockLocked").detail("Expecting", id).detail("Lock", BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()));
throw database_locked();
}
}
tr->atomicOp(databaseLockedKey,
BinaryWriter::toValue(id, Unversioned())
.withPrefix(LiteralStringRef("0123456789"))
.withSuffix(LiteralStringRef("\x00\x00\x00\x00")),
MutationRef::SetVersionstampedValue);
tr->addWriteConflictRange(normalKeys);
return Void();
}
ACTOR Future<Void> lockDatabase(Reference<ReadYourWritesTransaction> tr, UID id) {
tr->setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr->get(databaseLockedKey));
if (val.present()) {
if (BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()) == id) {
return Void();
} else {
//TraceEvent("DBA_LockLocked").detail("Expecting", id).detail("Lock", BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()));
throw database_locked();
}
}
tr->atomicOp(databaseLockedKey,
BinaryWriter::toValue(id, Unversioned())
.withPrefix(LiteralStringRef("0123456789"))
.withSuffix(LiteralStringRef("\x00\x00\x00\x00")),
MutationRef::SetVersionstampedValue);
tr->addWriteConflictRange(normalKeys);
return Void();
}
ACTOR Future<Void> lockDatabase(Database cx, UID id) {
state Transaction tr(cx);
loop {
try {
wait(lockDatabase(&tr, id));
wait(tr.commit());
return Void();
} catch (Error& e) {
if (e.code() == error_code_database_locked)
throw e;
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> unlockDatabase(Transaction* tr, UID id) {
tr->setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr->get(databaseLockedKey));
if (!val.present())
return Void();
if (val.present() && BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()) != id) {
//TraceEvent("DBA_UnlockLocked").detail("Expecting", id).detail("Lock", BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()));
throw database_locked();
}
tr->clear(singleKeyRange(databaseLockedKey));
return Void();
}
ACTOR Future<Void> unlockDatabase(Reference<ReadYourWritesTransaction> tr, UID id) {
tr->setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr->get(databaseLockedKey));
if (!val.present())
return Void();
if (val.present() && BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()) != id) {
//TraceEvent("DBA_UnlockLocked").detail("Expecting", id).detail("Lock", BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()));
throw database_locked();
}
tr->clear(singleKeyRange(databaseLockedKey));
return Void();
}
ACTOR Future<Void> unlockDatabase(Database cx, UID id) {
state Transaction tr(cx);
loop {
try {
wait(unlockDatabase(&tr, id));
wait(tr.commit());
return Void();
} catch (Error& e) {
if (e.code() == error_code_database_locked)
throw e;
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> checkDatabaseLock(Transaction* tr, UID id) {
tr->setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr->get(databaseLockedKey));
if (val.present() && BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()) != id) {
//TraceEvent("DBA_CheckLocked").detail("Expecting", id).detail("Lock", BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned())).backtrace();
throw database_locked();
}
return Void();
}
ACTOR Future<Void> checkDatabaseLock(Reference<ReadYourWritesTransaction> tr, UID id) {
tr->setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr->setOption(FDBTransactionOptions::LOCK_AWARE);
Optional<Value> val = wait(tr->get(databaseLockedKey));
if (val.present() && BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned()) != id) {
//TraceEvent("DBA_CheckLocked").detail("Expecting", id).detail("Lock", BinaryReader::fromStringRef<UID>(val.get().substr(10), Unversioned())).backtrace();
throw database_locked();
}
return Void();
}
ACTOR Future<Void> advanceVersion(Database cx, Version v) {
state Transaction tr(cx);
loop {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
try {
Version rv = wait(tr.getReadVersion());
if (rv <= v) {
tr.set(minRequiredCommitVersionKey, BinaryWriter::toValue(v + 1, Unversioned()));
wait(tr.commit());
} else {
printf("Current read version is %ld\n", rv);
return Void();
}
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> forceRecovery(Reference<ClusterConnectionFile> clusterFile, Key dcId) {
state Reference<AsyncVar<Optional<ClusterInterface>>> clusterInterface(new AsyncVar<Optional<ClusterInterface>>);
state Future<Void> leaderMon = monitorLeader<ClusterInterface>(clusterFile, clusterInterface);
loop {
choose {
when(wait(clusterInterface->get().present()
? brokenPromiseToNever(
clusterInterface->get().get().forceRecovery.getReply(ForceRecoveryRequest(dcId)))
: Never())) {
return Void();
}
when(wait(clusterInterface->onChange())) {}
}
}
}
ACTOR Future<Void> waitForPrimaryDC(Database cx, StringRef dcId) {
state ReadYourWritesTransaction tr(cx);
loop {
try {
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
Optional<Value> res = wait(tr.get(primaryDatacenterKey));
if (res.present() && res.get() == dcId) {
return Void();
}
state Future<Void> watchFuture = tr.watch(primaryDatacenterKey);
wait(tr.commit());
wait(watchFuture);
tr.reset();
} catch (Error& e) {
wait(tr.onError(e));
}
}
}
ACTOR Future<Void> changeCachedRange(Database cx, KeyRangeRef range, bool add) {
state ReadYourWritesTransaction tr(cx);
state KeyRange sysRange = KeyRangeRef(storageCacheKey(range.begin), storageCacheKey(range.end));
state KeyRange sysRangeClear = KeyRangeRef(storageCacheKey(range.begin), keyAfter(storageCacheKey(range.end)));
state KeyRange privateRange = KeyRangeRef(cacheKeysKey(0, range.begin), cacheKeysKey(0, range.end));
state Value trueValue = storageCacheValue(std::vector<uint16_t>{ 0 });
state Value falseValue = storageCacheValue(std::vector<uint16_t>{});
loop {
tr.setOption(FDBTransactionOptions::LOCK_AWARE);
tr.setOption(FDBTransactionOptions::ACCESS_SYSTEM_KEYS);
try {
tr.clear(sysRangeClear);
tr.clear(privateRange);
tr.addReadConflictRange(privateRange);
RangeResult previous = wait(tr.getRange(KeyRangeRef(storageCachePrefix, sysRange.begin), 1, true));
bool prevIsCached = false;
if (!previous.empty()) {
std::vector<uint16_t> prevVal;
decodeStorageCacheValue(previous[0].value, prevVal);
prevIsCached = !prevVal.empty();
}
if (prevIsCached && !add) {
// we need to uncache from here
tr.set(sysRange.begin, falseValue);
tr.set(privateRange.begin, serverKeysFalse);
} else if (!prevIsCached && add) {
// we need to cache, starting from here
tr.set(sysRange.begin, trueValue);
tr.set(privateRange.begin, serverKeysTrue);
}
RangeResult after = wait(tr.getRange(KeyRangeRef(sysRange.end, storageCacheKeys.end), 1, false));
bool afterIsCached = false;
if (!after.empty()) {
std::vector<uint16_t> afterVal;
decodeStorageCacheValue(after[0].value, afterVal);
afterIsCached = afterVal.empty();
}
if (afterIsCached && !add) {
tr.set(sysRange.end, trueValue);
tr.set(privateRange.end, serverKeysTrue);
} else if (!afterIsCached && add) {
tr.set(sysRange.end, falseValue);
tr.set(privateRange.end, serverKeysFalse);
}
wait(tr.commit());
return Void();
} catch (Error& e) {
state Error err = e;
wait(tr.onError(err));
TraceEvent(SevDebug, "ChangeCachedRangeError").error(err);
}
}
}
Future<Void> addCachedRange(const Database& cx, KeyRangeRef range) {
return changeCachedRange(cx, range, true);
}
Future<Void> removeCachedRange(const Database& cx, KeyRangeRef range) {
return changeCachedRange(cx, range, false);
}
json_spirit::Value_type normJSONType(json_spirit::Value_type type) {
if (type == json_spirit::int_type)
return json_spirit::real_type;
return type;
}
void schemaCoverage(std::string const& spath, bool covered) {
static std::map<bool, std::set<std::string>> coveredSchemaPaths;
if (coveredSchemaPaths[covered].insert(spath).second) {
TraceEvent ev(SevInfo, "CodeCoverage");
ev.detail("File", "documentation/StatusSchema.json/" + spath).detail("Line", 0);
if (!covered)
ev.detail("Covered", 0);
}
}
bool schemaMatch(json_spirit::mValue const& schemaValue,
json_spirit::mValue const& resultValue,
std::string& errorStr,
Severity sev,
bool checkCoverage,
std::string path,
std::string schemaPath) {
// Returns true if everything in `result` is permitted by `schema`
bool ok = true;
try {
if (normJSONType(schemaValue.type()) != normJSONType(resultValue.type())) {
errorStr += format("ERROR: Incorrect value type for key `%s'\n", path.c_str());
TraceEvent(sev, "SchemaMismatch")
.detail("Path", path)
.detail("SchemaType", schemaValue.type())
.detail("ValueType", resultValue.type());
return false;
}
if (resultValue.type() == json_spirit::obj_type) {
auto& result = resultValue.get_obj();
auto& schema = schemaValue.get_obj();
for (auto& rkv : result) {
auto& key = rkv.first;
auto& rv = rkv.second;
std::string kpath = path + "." + key;
std::string spath = schemaPath + "." + key;
if (checkCoverage) {
schemaCoverage(spath);
}
if (!schema.count(key)) {
errorStr += format("ERROR: Unknown key `%s'\n", kpath.c_str());
TraceEvent(sev, "SchemaMismatch").detail("Path", kpath).detail("SchemaPath", spath);
ok = false;
continue;
}
auto& sv = schema.at(key);
if (sv.type() == json_spirit::obj_type && sv.get_obj().count("$enum")) {
auto& enum_values = sv.get_obj().at("$enum").get_array();
bool any_match = false;
for (auto& enum_item : enum_values)
if (enum_item == rv) {
any_match = true;
if (checkCoverage) {
schemaCoverage(spath + ".$enum." + enum_item.get_str());
}
break;
}
if (!any_match) {
errorStr += format("ERROR: Unknown value `%s' for key `%s'\n",
json_spirit::write_string(rv).c_str(),
kpath.c_str());
TraceEvent(sev, "SchemaMismatch")
.detail("Path", kpath)
.detail("SchemaEnumItems", enum_values.size())
.detail("Value", json_spirit::write_string(rv));
if (checkCoverage) {
schemaCoverage(spath + ".$enum." + json_spirit::write_string(rv));
}
ok = false;
}
} else if (sv.type() == json_spirit::obj_type && sv.get_obj().count("$map")) {
if (rv.type() != json_spirit::obj_type) {
errorStr += format("ERROR: Expected an object as the value for key `%s'\n", kpath.c_str());
TraceEvent(sev, "SchemaMismatch")
.detail("Path", kpath)
.detail("SchemaType", sv.type())
.detail("ValueType", rv.type());
ok = false;
continue;
}
if (sv.get_obj().at("$map").type() != json_spirit::obj_type) {
continue;
}
auto& schemaVal = sv.get_obj().at("$map");
auto& valueObj = rv.get_obj();
if (checkCoverage) {
schemaCoverage(spath + ".$map");
}
for (auto& valuePair : valueObj) {
auto vpath = kpath + "[" + valuePair.first + "]";
auto upath = spath + ".$map";
if (valuePair.second.type() != json_spirit::obj_type) {
errorStr += format("ERROR: Expected an object for `%s'\n", vpath.c_str());
TraceEvent(sev, "SchemaMismatch")
.detail("Path", vpath)
.detail("ValueType", valuePair.second.type());
ok = false;
continue;
}
if (!schemaMatch(schemaVal, valuePair.second, errorStr, sev, checkCoverage, vpath, upath)) {
ok = false;
}
}
} else {
if (!schemaMatch(sv, rv, errorStr, sev, checkCoverage, kpath, spath)) {
ok = false;
}
}
}
} else if (resultValue.type() == json_spirit::array_type) {
auto& valueArray = resultValue.get_array();
auto& schemaArray = schemaValue.get_array();
if (!schemaArray.size()) {
// An empty schema array means that the value array is required to be empty
if (valueArray.size()) {
errorStr += format("ERROR: Expected an empty array for key `%s'\n", path.c_str());
TraceEvent(sev, "SchemaMismatch")
.detail("Path", path)
.detail("SchemaSize", schemaArray.size())
.detail("ValueSize", valueArray.size());
return false;
}
} else if (schemaArray.size() == 1) {
// A one item schema array means that all items in the value must match the first item in the schema
int index = 0;
for (auto& valueItem : valueArray) {
if (!schemaMatch(schemaArray[0],
valueItem,
errorStr,
sev,
checkCoverage,
path + format("[%d]", index),
schemaPath + "[0]")) {
ok = false;
}
index++;
}
} else {
ASSERT(false); // Schema doesn't make sense
}
}
return ok;
} catch (std::exception& e) {
TraceEvent(SevError, "SchemaMatchException")
.detail("What", e.what())
.detail("Path", path)
.detail("SchemaPath", schemaPath);
throw unknown_error();
}
}
TEST_CASE("/ManagementAPI/AutoQuorumChange/checkLocality") {
wait(Future<Void>(Void()));
std::vector<ProcessData> workers;
std::vector<NetworkAddress> chosen;
std::set<AddressExclusion> excluded;
AutoQuorumChange change(5);
for (int i = 0; i < 10; i++) {
ProcessData data;
auto dataCenter = std::to_string(i / 4 % 2);
auto dataHall = dataCenter + std::to_string(i / 2 % 2);
auto rack = dataHall + std::to_string(i % 2);
auto machineId = rack + std::to_string(i);
data.locality.set(LiteralStringRef("dcid"), StringRef(dataCenter));
data.locality.set(LiteralStringRef("data_hall"), StringRef(dataHall));
data.locality.set(LiteralStringRef("rack"), StringRef(rack));
data.locality.set(LiteralStringRef("zoneid"), StringRef(rack));
data.locality.set(LiteralStringRef("machineid"), StringRef(machineId));
data.address.ip = IPAddress(i);
if (g_network->isSimulated()) {
g_simulator.newProcess("TestCoordinator",
data.address.ip,
data.address.port,
false,
1,
data.locality,
ProcessClass(ProcessClass::CoordinatorClass, ProcessClass::CommandLineSource),
"",
"",
currentProtocolVersion);
}
workers.push_back(data);
}
auto noAssignIndex = deterministicRandom()->randomInt(0, workers.size());
workers[noAssignIndex].processClass._class = ProcessClass::CoordinatorClass;
change.addDesiredWorkers(chosen, workers, 5, excluded);
std::map<StringRef, std::set<StringRef>> chosenValues;
ASSERT(chosen.size() == 5);
std::vector<StringRef> fields({ LiteralStringRef("dcid"),
LiteralStringRef("data_hall"),
LiteralStringRef("zoneid"),
LiteralStringRef("machineid") });
for (auto worker = chosen.begin(); worker != chosen.end(); worker++) {
ASSERT(worker->ip.toV4() < workers.size());
LocalityData data = workers[worker->ip.toV4()].locality;
for (auto field = fields.begin(); field != fields.end(); field++) {
chosenValues[*field].insert(data.get(*field).get());
}
}
ASSERT(chosenValues[LiteralStringRef("dcid")].size() == 2);
ASSERT(chosenValues[LiteralStringRef("data_hall")].size() == 4);
ASSERT(chosenValues[LiteralStringRef("zoneid")].size() == 5);
ASSERT(chosenValues[LiteralStringRef("machineid")].size() == 5);
ASSERT(std::find(chosen.begin(), chosen.end(), workers[noAssignIndex].address) != chosen.end());
return Void();
}
Reference in New Issue View Git Blame Copy Permalink