Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
68.94% |
91 / 132 |
|
38.46% |
5 / 13 |
CRAP | |
0.00% |
0 / 1 |
| LBFactoryMulti | |
68.94% |
91 / 132 |
|
38.46% |
5 / 13 |
78.02 | |
0.00% |
0 / 1 |
| __construct | |
73.91% |
17 / 23 |
|
0.00% |
0 / 1 |
5.44 | |||
| newMainLB | |
87.50% |
14 / 16 |
|
0.00% |
0 / 1 |
3.02 | |||
| resolveDomainInstance | |
69.23% |
9 / 13 |
|
0.00% |
0 / 1 |
7.05 | |||
| getMainLB | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
2 | |||
| newExternalLB | |
91.67% |
11 / 12 |
|
0.00% |
0 / 1 |
2.00 | |||
| getExternalLB | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
2 | |||
| getAllMainLBs | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
12 | |||
| getAllExternalLBs | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
6 | |||
| getLBsForOwner | |
75.00% |
3 / 4 |
|
0.00% |
0 / 1 |
3.14 | |||
| newLoadBalancer | |
100.00% |
11 / 11 |
|
100.00% |
1 / 1 |
1 | |||
| makeServerConfigArrays | |
100.00% |
13 / 13 |
|
100.00% |
1 / 1 |
3 | |||
| getSectionFromDatabase | |
100.00% |
3 / 3 |
|
100.00% |
1 / 1 |
1 | |||
| reconfigure | |
0.00% |
0 / 18 |
|
0.00% |
0 / 1 |
20 | |||
| 1 | <?php |
| 2 | /** |
| 3 | * @license GPL-2.0-or-later |
| 4 | * @file |
| 5 | */ |
| 6 | namespace Wikimedia\Rdbms; |
| 7 | |
| 8 | use InvalidArgumentException; |
| 9 | use LogicException; |
| 10 | use UnexpectedValueException; |
| 11 | |
| 12 | /** |
| 13 | * LoadBalancer manager for sites with several "main" database clusters |
| 14 | * |
| 15 | * Each database cluster consists of a "primary" server and any number of replica servers, |
| 16 | * all of which converge, as soon as possible, to contain the same schemas and records. If |
| 17 | * a replication topology has multiple primaries, then the "primary" is merely the preferred |
| 18 | * co-primary for the current context (e.g. datacenter). |
| 19 | * |
| 20 | * For single-primary topologies, the schemas and records of the primary define the "dataset". |
| 21 | * For multiple-primary topologies, the "dataset" is the convergent result of applying/merging |
| 22 | * all committed events (regardless of the co-primary they originated on); it possible that no |
| 23 | * co-primary has yet converged upon this state at any given time (especially when there are |
| 24 | * frequent writes and co-primaries are geographically distant). |
| 25 | * |
| 26 | * A "main" cluster contain a "main" dataset, which consists of data that is compact, highly |
| 27 | * relational (e.g. read by JOIN queries), and essential to one or more sites. The "external" |
| 28 | * clusters each store an "external" dataset, which consists of data that is non-relational |
| 29 | * (e.g. key/value pairs), self-contained (e.g. JOIN queries and transactions thereof never |
| 30 | * involve a main dataset), or too bulky to reside in a main dataset (e.g. text blobs). |
| 31 | * |
| 32 | * The class allows for large site farms to split up their data in the following ways: |
| 33 | * - Vertically shard compact site-specific data by site (e.g. page/comment metadata) |
| 34 | * - Vertically shard compact global data by module (e.g. account/notification data) |
| 35 | * - Horizontally shard any bulk data by blob key (e.g. page/comment content blobs) |
| 36 | * |
| 37 | * @ingroup Database |
| 38 | */ |
| 39 | class LBFactoryMulti extends LBFactory { |
| 40 | /** @var array<string,ILoadBalancerForOwner> Map of (main section => tracked LoadBalancer) */ |
| 41 | private $mainLBs = []; |
| 42 | /** @var array<string,ILoadBalancerForOwner> Map of (external cluster => tracked LoadBalancer) */ |
| 43 | private $externalLBs = []; |
| 44 | |
| 45 | /** @var string[] Map of (server name => IP address) */ |
| 46 | private $hostsByServerName; |
| 47 | /** @var string[] Map of (database name => main section) */ |
| 48 | private $sectionsByDB; |
| 49 | /** @var int[][] Map of (main section => server name => load ratio) */ |
| 50 | private $loadsBySection; |
| 51 | /** @var int[][] Map of (external cluster => server name => load ratio) */ |
| 52 | private $externalLoadsByCluster; |
| 53 | /** @var array Server config map ("host", "serverName", and "load" ignored) */ |
| 54 | private $serverTemplate; |
| 55 | /** @var array Server config map overriding "serverTemplate" for all external servers */ |
| 56 | private $externalTemplateOverrides; |
| 57 | /** @var array[] Map of (main section => server config map overrides) */ |
| 58 | private $templateOverridesBySection; |
| 59 | /** @var array[] Map of (external cluster => server config map overrides) */ |
| 60 | private $templateOverridesByCluster; |
| 61 | /** @var array Server config override map for all main/external primary DB servers */ |
| 62 | private $masterTemplateOverrides; |
| 63 | /** @var array[] Map of (server name => server config map overrides) for all servers */ |
| 64 | private $templateOverridesByServer; |
| 65 | /** @var string[]|bool[] A map of (main section => read-only message) */ |
| 66 | private $readOnlyBySection; |
| 67 | /** @var array Configuration for the LoadMonitor to use within LoadBalancer instances */ |
| 68 | private $loadMonitorConfig; |
| 69 | /** @var DatabaseDomain[] Map of (domain ID => domain instance) */ |
| 70 | private $nonLocalDomainCache = []; |
| 71 | |
| 72 | /** |
| 73 | * Template override precedence (highest => lowest): |
| 74 | * - templateOverridesByServer |
| 75 | * - masterTemplateOverrides |
| 76 | * - templateOverridesBySection/templateOverridesByCluster |
| 77 | * - externalTemplateOverrides |
| 78 | * - serverTemplate |
| 79 | * Overrides only work on top level keys (so nested values will not be merged). |
| 80 | * |
| 81 | * Server config maps should be of the format Database::factory() requires. |
| 82 | * Additionally, a 'max lag' key should also be set on server maps, indicating how stale the |
| 83 | * data can be before the load balancer tries to avoid using it. The map can have 'is static' |
| 84 | * set to disable blocking replication sync checks (intended for archive servers with |
| 85 | * unchanging data). |
| 86 | * |
| 87 | * @see LBFactory::__construct() |
| 88 | * @param array $conf Additional parameters include: |
| 89 | * - hostsByName: map of (server name => IP address). [optional] |
| 90 | * - sectionsByDB: map of (database => main section). The database name "DEFAULT" is |
| 91 | * interpreted as a catch-all for all databases not otherwise mentioned. If no section |
| 92 | * name is specified for "DEFAULT", then the catch-all section is assumed to be named |
| 93 | * "DEFAULT". [optional] |
| 94 | * - sectionLoads: map of (main section => server name => load ratio); the first host |
| 95 | * listed in each section is the primary DB server for that section. [optional] |
| 96 | * - externalLoads: map of (cluster => server name => load ratio) map. [optional] |
| 97 | * - serverTemplate: server config map for Database::factory(). |
| 98 | * Note that "host", "serverName" and "load" entries will be overridden by "hostsByName". [optional] |
| 99 | * - externalTemplateOverrides: server config map overrides for external stores; |
| 100 | * respects the override precedence described above. [optional] |
| 101 | * - templateOverridesBySection: map of (main section => server config map overrides); |
| 102 | * respects the override precedence described above. [optional] |
| 103 | * - templateOverridesByCluster: map of (external cluster => server config map overrides); |
| 104 | * respects the override precedence described above. [optional] |
| 105 | * - masterTemplateOverrides: server config map overrides for masters; |
| 106 | * respects the override precedence described above. [optional] |
| 107 | * - templateOverridesByServer: map of (server name => server config map overrides); |
| 108 | * respects the override precedence described above and applies to both core |
| 109 | * and external storage. [optional] |
| 110 | * - loadMonitor: LoadMonitor::__construct() parameters with "class" field. [optional] |
| 111 | * - readOnlyBySection: map of (main section => message text or false). |
| 112 | * String values make sections read only, whereas anything else does not |
| 113 | * restrict read/write mode. [optional] |
| 114 | * - configCallback: A callback that returns a conf array that can be passed to |
| 115 | * the reconfigure() method. This will be used to autoReconfigure() to load |
| 116 | * any updated configuration. |
| 117 | */ |
| 118 | public function __construct( array $conf ) { |
| 119 | parent::__construct( $conf ); |
| 120 | |
| 121 | $this->hostsByServerName = $conf['hostsByName'] ?? []; |
| 122 | $this->sectionsByDB = $conf['sectionsByDB']; |
| 123 | $this->sectionsByDB += [ self::CLUSTER_MAIN_DEFAULT => self::CLUSTER_MAIN_DEFAULT ]; |
| 124 | $this->loadsBySection = $conf['sectionLoads'] ?? []; |
| 125 | $this->externalLoadsByCluster = $conf['externalLoads'] ?? []; |
| 126 | $this->serverTemplate = $conf['serverTemplate'] ?? []; |
| 127 | $this->externalTemplateOverrides = $conf['externalTemplateOverrides'] ?? []; |
| 128 | $this->templateOverridesBySection = $conf['templateOverridesBySection'] ?? []; |
| 129 | $this->templateOverridesByCluster = $conf['templateOverridesByCluster'] ?? []; |
| 130 | $this->masterTemplateOverrides = $conf['masterTemplateOverrides'] ?? []; |
| 131 | $this->templateOverridesByServer = $conf['templateOverridesByServer'] ?? []; |
| 132 | $this->readOnlyBySection = $conf['readOnlyBySection'] ?? []; |
| 133 | |
| 134 | if ( isset( $conf['loadMonitor'] ) ) { |
| 135 | $this->loadMonitorConfig = $conf['loadMonitor']; |
| 136 | } elseif ( isset( $conf['loadMonitorClass'] ) ) { // b/c |
| 137 | $this->loadMonitorConfig = [ 'class' => $conf['loadMonitorClass'] ]; |
| 138 | } else { |
| 139 | $this->loadMonitorConfig = [ 'class' => LoadMonitor::class ]; |
| 140 | } |
| 141 | |
| 142 | foreach ( $this->externalLoadsByCluster as $cluster => $_ ) { |
| 143 | if ( isset( $this->loadsBySection[$cluster] ) ) { |
| 144 | throw new LogicException( |
| 145 | "External cluster '$cluster' has the same name as a main section/cluster" |
| 146 | ); |
| 147 | } |
| 148 | } |
| 149 | } |
| 150 | |
| 151 | /** @inheritDoc */ |
| 152 | public function newMainLB( $domain = false ): ILoadBalancerForOwner { |
| 153 | $domainInstance = $this->resolveDomainInstance( $domain ); |
| 154 | $database = $domainInstance->getDatabase(); |
| 155 | $section = $this->getSectionFromDatabase( $database ); |
| 156 | |
| 157 | if ( !isset( $this->loadsBySection[$section] ) ) { |
| 158 | throw new UnexpectedValueException( "Section '$section' has no hosts defined." ); |
| 159 | } |
| 160 | |
| 161 | return $this->newLoadBalancer( |
| 162 | $section, |
| 163 | array_merge( |
| 164 | $this->serverTemplate, |
| 165 | $this->templateOverridesBySection[$section] ?? [] |
| 166 | ), |
| 167 | $this->loadsBySection[$section], |
| 168 | // Use the LB-specific read-only reason if everything isn't already read-only |
| 169 | is_string( $this->readOnlyReason ) |
| 170 | ? $this->readOnlyReason |
| 171 | : ( $this->readOnlyBySection[$section] ?? false ) |
| 172 | ); |
| 173 | } |
| 174 | |
| 175 | /** |
| 176 | * @param DatabaseDomain|string|false $domain |
| 177 | * @return DatabaseDomain |
| 178 | */ |
| 179 | private function resolveDomainInstance( $domain ) { |
| 180 | if ( $domain instanceof DatabaseDomain ) { |
| 181 | return $domain; // already a domain instance |
| 182 | } elseif ( $domain === false || $domain === $this->localDomain->getId() ) { |
| 183 | return $this->localDomain; |
| 184 | } elseif ( isset( $this->domainAliases[$domain] ) ) { |
| 185 | // This array acts as both the original map and as instance cache. |
| 186 | // Instances pass-through DatabaseDomain::newFromId as-is. |
| 187 | $this->domainAliases[$domain] = |
| 188 | DatabaseDomain::newFromId( $this->domainAliases[$domain] ); |
| 189 | |
| 190 | return $this->domainAliases[$domain]; |
| 191 | } |
| 192 | |
| 193 | $cachedDomain = $this->nonLocalDomainCache[$domain] ?? null; |
| 194 | if ( $cachedDomain === null ) { |
| 195 | $cachedDomain = DatabaseDomain::newFromId( $domain ); |
| 196 | $this->nonLocalDomainCache = [ $domain => $cachedDomain ]; |
| 197 | } |
| 198 | |
| 199 | return $cachedDomain; |
| 200 | } |
| 201 | |
| 202 | /** @inheritDoc */ |
| 203 | public function getMainLB( $domain = false ): ILoadBalancer { |
| 204 | $domainInstance = $this->resolveDomainInstance( $domain ); |
| 205 | $section = $this->getSectionFromDatabase( $domainInstance->getDatabase() ); |
| 206 | |
| 207 | if ( !isset( $this->mainLBs[$section] ) ) { |
| 208 | $this->mainLBs[$section] = $this->newMainLB( $domain ); |
| 209 | } |
| 210 | |
| 211 | return $this->mainLBs[$section]; |
| 212 | } |
| 213 | |
| 214 | /** @inheritDoc */ |
| 215 | public function newExternalLB( $cluster ): ILoadBalancerForOwner { |
| 216 | if ( !isset( $this->externalLoadsByCluster[$cluster] ) ) { |
| 217 | throw new InvalidArgumentException( "Unknown cluster '$cluster'" ); |
| 218 | } |
| 219 | return $this->newLoadBalancer( |
| 220 | $cluster, |
| 221 | array_merge( |
| 222 | $this->serverTemplate, |
| 223 | $this->externalTemplateOverrides, |
| 224 | $this->templateOverridesByCluster[$cluster] ?? [] |
| 225 | ), |
| 226 | $this->externalLoadsByCluster[$cluster], |
| 227 | $this->readOnlyReason |
| 228 | ); |
| 229 | } |
| 230 | |
| 231 | /** @inheritDoc */ |
| 232 | public function getExternalLB( $cluster ): ILoadBalancer { |
| 233 | if ( !isset( $this->externalLBs[$cluster] ) ) { |
| 234 | $this->externalLBs[$cluster] = $this->newExternalLB( |
| 235 | $cluster |
| 236 | ); |
| 237 | } |
| 238 | |
| 239 | return $this->externalLBs[$cluster]; |
| 240 | } |
| 241 | |
| 242 | public function getAllMainLBs(): array { |
| 243 | $lbs = []; |
| 244 | foreach ( $this->sectionsByDB as $db => $section ) { |
| 245 | if ( !isset( $lbs[$section] ) ) { |
| 246 | $lbs[$section] = $this->getMainLB( $db ); |
| 247 | } |
| 248 | } |
| 249 | |
| 250 | return $lbs; |
| 251 | } |
| 252 | |
| 253 | public function getAllExternalLBs(): array { |
| 254 | $lbs = []; |
| 255 | foreach ( $this->externalLoadsByCluster as $cluster => $unused ) { |
| 256 | $lbs[$cluster] = $this->getExternalLB( $cluster ); |
| 257 | } |
| 258 | |
| 259 | return $lbs; |
| 260 | } |
| 261 | |
| 262 | /** @inheritDoc */ |
| 263 | protected function getLBsForOwner() { |
| 264 | foreach ( $this->mainLBs as $lb ) { |
| 265 | yield $lb; |
| 266 | } |
| 267 | foreach ( $this->externalLBs as $lb ) { |
| 268 | yield $lb; |
| 269 | } |
| 270 | } |
| 271 | |
| 272 | /** |
| 273 | * Make a new load balancer object based on template and load array |
| 274 | * |
| 275 | * @param string $clusterName |
| 276 | * @param array $serverTemplate |
| 277 | * @param array $loads |
| 278 | * @param string|false $readOnlyReason |
| 279 | * @return LoadBalancer |
| 280 | */ |
| 281 | private function newLoadBalancer( |
| 282 | string $clusterName, |
| 283 | array $serverTemplate, |
| 284 | array $loads, |
| 285 | $readOnlyReason |
| 286 | ) { |
| 287 | $lb = new LoadBalancer( array_merge( |
| 288 | $this->baseLoadBalancerParams(), |
| 289 | [ |
| 290 | 'servers' => $this->makeServerConfigArrays( $serverTemplate, $loads ), |
| 291 | 'loadMonitor' => $this->loadMonitorConfig, |
| 292 | 'readOnlyReason' => $readOnlyReason, |
| 293 | 'clusterName' => $clusterName |
| 294 | ] |
| 295 | ) ); |
| 296 | $this->initLoadBalancer( $lb ); |
| 297 | |
| 298 | return $lb; |
| 299 | } |
| 300 | |
| 301 | /** |
| 302 | * Make a server array as expected by LoadBalancer::__construct() |
| 303 | * |
| 304 | * @param array $serverTemplate Server config map |
| 305 | * @param int[] $loads Map of (server name => load) |
| 306 | * @return array[] List of server config maps |
| 307 | */ |
| 308 | private function makeServerConfigArrays( array $serverTemplate, array $loads ) { |
| 309 | // Get the ordered map of (server name => load); the primary DB server is first |
| 310 | $servers = []; |
| 311 | foreach ( $loads as $serverName => $load ) { |
| 312 | $servers[] = array_merge( |
| 313 | $serverTemplate, |
| 314 | $servers ? [] : $this->masterTemplateOverrides, |
| 315 | $this->templateOverridesByServer[$serverName] ?? [], |
| 316 | [ |
| 317 | 'host' => $this->hostsByServerName[$serverName] ?? $serverName, |
| 318 | 'serverName' => $serverName, |
| 319 | 'load' => $load, |
| 320 | ] |
| 321 | ); |
| 322 | } |
| 323 | |
| 324 | return $servers; |
| 325 | } |
| 326 | |
| 327 | /** |
| 328 | * @param string $database |
| 329 | * @return string Main section name |
| 330 | */ |
| 331 | private function getSectionFromDatabase( $database ) { |
| 332 | return $this->sectionsByDB[$database] |
| 333 | ?? $this->sectionsByDB[self::CLUSTER_MAIN_DEFAULT] |
| 334 | ?? self::CLUSTER_MAIN_DEFAULT; |
| 335 | } |
| 336 | |
| 337 | public function reconfigure( array $conf ): void { |
| 338 | if ( !$conf ) { |
| 339 | return; |
| 340 | } |
| 341 | |
| 342 | foreach ( $this->mainLBs as $lb ) { |
| 343 | // Approximate what LBFactoryMulti::__construct does (T346365) |
| 344 | $config = [ |
| 345 | 'servers' => $this->makeServerConfigArrays( |
| 346 | $conf['serverTemplate'] ?? [], |
| 347 | $conf['sectionLoads'][$lb->getClusterName()] |
| 348 | ) |
| 349 | ]; |
| 350 | $lb->reconfigure( $config ); |
| 351 | |
| 352 | } |
| 353 | foreach ( $this->externalLBs as $lb ) { |
| 354 | $config = [ |
| 355 | 'servers' => $this->makeServerConfigArrays( |
| 356 | $conf['serverTemplate'] ?? [], |
| 357 | $conf['externalLoads'][$lb->getClusterName()] |
| 358 | ) |
| 359 | ]; |
| 360 | $lb->reconfigure( $config ); |
| 361 | } |
| 362 | } |
| 363 | } |