|
120 | 120 | severity: 'warning',
|
121 | 121 | },
|
122 | 122 | },
|
| 123 | + { |
| 124 | + alert: 'PostgresHasTooManyRollbacks', |
| 125 | + annotations: { |
| 126 | + description: 'PostgreSQL has too many rollbacks on {{ $labels.cluster }} for database {{ $labels.datname }} with a value of {{ $value }}', |
| 127 | + summary: 'PostgreSQL has too many rollbacks.', |
| 128 | + }, |
| 129 | + expr: ||| |
| 130 | + avg without(pod, instance) |
| 131 | + (rate(pg_stat_database_xact_rollback{db_name!~"template.*|^$"}[5m]) / |
| 132 | + (rate(pg_stat_database_xact_commit{db_name!~"template.*|^$"}[5m])+ rate(pg_stat_database_xact_rollback{db_name!~"template.*|^$"}[5m]))) > 0.10 |
| 133 | + ||| % $._config, |
| 134 | + 'for': '5m', |
| 135 | + labels: { |
| 136 | + severity: 'warning', |
| 137 | + }, |
| 138 | + }, |
| 139 | + { |
| 140 | + alert: 'PostgresHasHighDeadLocks', |
| 141 | + annotations: { |
| 142 | + description: 'PostgreSQL has too high deadlocks on {{ $labels.cluster }} for database {{ $labels.datname }} with a value of {{ $value }}', |
| 143 | + summary: 'PostgreSQL has high number of deadlocks.', |
| 144 | + }, |
| 145 | + expr: ||| |
| 146 | + max without(pod, instance) (rate(pg_stat_database_deadlocks{db_name!~"template.*|^$"}[5m]) * 60) > 5 |
| 147 | + ||| % $._config, |
| 148 | + 'for': '5m', |
| 149 | + labels: { |
| 150 | + severity: 'warning', |
| 151 | + }, |
| 152 | + }, |
| 153 | + { |
| 154 | + alert: 'PostgresAcquiredTooManyLocks', |
| 155 | + annotations: { |
| 156 | + description: 'PostgreSQL has acquired too many locks on {{ $labels.cluster }} for database {{ $labels.datname }} with a value of {{ $value }}', |
| 157 | + summary: 'PostgreSQL has high number of acquired locks.', |
| 158 | + }, |
| 159 | + expr: ||| |
| 160 | + max by( server, job, db_name, asserts_env, asserts_site, namespace) ((pg_locks_count{db_name!~"template.*|^$"}) / |
| 161 | + on(instance, asserts_env, asserts_site, namespace) group_left(server) (pg_settings_max_locks_per_transaction{} * pg_settings_max_connections{})) > 0.20 |
| 162 | + ||| % $._config, |
| 163 | + 'for': '5m', |
| 164 | + labels: { |
| 165 | + severity: 'warning', |
| 166 | + }, |
| 167 | + }, |
| 168 | + { |
| 169 | + alert: 'PostgresXLOGConsumptionVeryLow', |
| 170 | + annotations: { |
| 171 | + description: 'PostgreSQL instance {{ $labels.instance }} has a very low XLOG consumption rate.', |
| 172 | + summary: 'PostgreSQL XLOG consumption is very low.', |
| 173 | + }, |
| 174 | + expr: 'rate(pg_xlog_position_bytes{asserts_env!=""}[5m]) < 200000', |
| 175 | + 'for': '5m', |
| 176 | + labels: { |
| 177 | + asserts_severity: 'critical', |
| 178 | + asserts_entity_type: 'Service', |
| 179 | + asserts_alert_category: 'failure', |
| 180 | + }, |
| 181 | + }, |
| 182 | + { |
| 183 | + alert: 'PostgresXLOGConsumptionVeryHigh', |
| 184 | + annotations: { |
| 185 | + description: '{{ $labels.instance }} is experiencing very high XLOG consumption rate, which might indicate excessive write operations.', |
| 186 | + summary: 'PostgreSQL very high XLOG consumption rate.', |
| 187 | + }, |
| 188 | + expr: 'rate(pg_xlog_position_bytes{asserts_env!=""}[2m]) > 36700160 and on (instance, asserts_env, asserts_site) (pg_replication_is_replica{asserts_env!=""} == 0)', |
| 189 | + 'for': '10m', |
| 190 | + labels: { |
| 191 | + asserts_severity: 'critical', |
| 192 | + asserts_entity_type: 'Service', |
| 193 | + asserts_alert_category: 'failure', |
| 194 | + }, |
| 195 | + }, |
| 196 | + { |
| 197 | + alert: 'PostgresReplicationStopped', |
| 198 | + annotations: { |
| 199 | + description: 'PostgreSQL instance {{ $labels.instance }} has stopped replication.', |
| 200 | + summary: 'PostgreSQL replication has stopped.', |
| 201 | + }, |
| 202 | + expr: 'pg_stat_replication_pg_xlog_location_diff{asserts_env!=""} != 0', |
| 203 | + 'for': '5m', |
| 204 | + labels: { |
| 205 | + asserts_severity: 'critical', |
| 206 | + asserts_entity_type: 'Service', |
| 207 | + asserts_alert_category: 'failure', |
| 208 | + }, |
| 209 | + }, |
| 210 | + { |
| 211 | + alert: 'PostgresReplicationLagging_More_1Hour', |
| 212 | + annotations: { |
| 213 | + description: '{{ $labels.instance }} replication lag exceeds 1 hour. Check for network issues or load imbalances.', |
| 214 | + summary: 'PostgreSQL replication lagging more than 1 hour.', |
| 215 | + }, |
| 216 | + expr: '(pg_replication_lag{asserts_env!=""} > 3600) and on (instance) (pg_replication_is_replica{asserts_env!=""} == 1)', |
| 217 | + 'for': '5m', |
| 218 | + labels: { |
| 219 | + asserts_severity: 'warning', |
| 220 | + asserts_entity_type: 'Service', |
| 221 | + asserts_alert_category: 'failure', |
| 222 | + }, |
| 223 | + }, |
| 224 | + { |
| 225 | + alert: 'PostgresReplicationLagBytesAreTooLarge', |
| 226 | + annotations: { |
| 227 | + description: '{{ $labels.instance }} replication lag in bytes is too large, which might indicate replication issues or network bottlenecks.', |
| 228 | + summary: 'PostgreSQL replication lag in bytes too large.', |
| 229 | + }, |
| 230 | + expr: '(pg_xlog_position_bytes{asserts_env!=""} and pg_replication_is_replica{asserts_env!=""} == 0) - on (job, service, asserts_env, asserts_site) group_right(instance) (pg_xlog_position_bytes{asserts_env!=""} and pg_replication_is_replica{asserts_env!=""} == 1) > 1e+09', |
| 231 | + 'for': '5m', |
| 232 | + labels: { |
| 233 | + asserts_severity: 'critical', |
| 234 | + asserts_entity_type: 'Service', |
| 235 | + asserts_alert_category: 'failure', |
| 236 | + }, |
| 237 | + }, |
| 238 | + { |
| 239 | + alert: 'PostgresHasReplicationSlotUsed', |
| 240 | + annotations: { |
| 241 | + description: '{{ $labels.instance }} has replication slots that are not used, which might lead to replication lag or data inconsistency.', |
| 242 | + summary: 'PostgreSQL has unused replication slots.', |
| 243 | + }, |
| 244 | + expr: 'pg_replication_slots_active{asserts_env!=""} == 0', |
| 245 | + 'for': '30m', |
| 246 | + labels: { |
| 247 | + asserts_severity: 'critical', |
| 248 | + asserts_entity_type: 'Service', |
| 249 | + asserts_alert_category: 'failure', |
| 250 | + }, |
| 251 | + }, |
| 252 | + { |
| 253 | + alert: 'PostgresReplicationIsStale', |
| 254 | + annotations: { |
| 255 | + description: '{{ $labels.instance }} replication slots have not been updated for a significant period, indicating potential issues with replication.', |
| 256 | + summary: 'PostgreSQL replication slots are stale.', |
| 257 | + }, |
| 258 | + expr: 'pg_replication_slots_xmin_age{asserts_env!="", slot_name =~ "^repmgr_slot_[0-9]+"} > 20000', |
| 259 | + 'for': '30m', |
| 260 | + labels: { |
| 261 | + asserts_severity: 'critical', |
| 262 | + asserts_entity_type: 'Service', |
| 263 | + asserts_alert_category: 'failure', |
| 264 | + }, |
| 265 | + }, |
| 266 | + { |
| 267 | + alert: 'PostgresReplicationRoleChanged', |
| 268 | + annotations: { |
| 269 | + description: '{{ $labels.instance }} replication role has changed. Verify if this is expected or if it indicates a failover.', |
| 270 | + summary: 'PostgreSQL replication role change detected.', |
| 271 | + }, |
| 272 | + expr: 'pg_replication_is_replica{asserts_env!=""} and changes(pg_replication_is_replica{asserts_env!=""}[1m]) > 0', |
| 273 | + labels: { |
| 274 | + asserts_severity: 'warning', |
| 275 | + asserts_entity_type: 'Service', |
| 276 | + asserts_alert_category: 'failure', |
| 277 | + }, |
| 278 | + }, |
| 279 | + { |
| 280 | + alert: 'PostgresHasExporterErrors', |
| 281 | + annotations: { |
| 282 | + description: '{{ $labels.instance }} exporter is experiencing errors. Verify exporter health and configuration.', |
| 283 | + summary: 'PostgreSQL exporter errors detected.', |
| 284 | + }, |
| 285 | + expr: 'pg_exporter_last_scrape_error{asserts_env!=""} > 0', |
| 286 | + 'for': '30m', |
| 287 | + labels: { |
| 288 | + asserts_severity: 'critical', |
| 289 | + asserts_entity_type: 'Service', |
| 290 | + asserts_alert_category: 'failure', |
| 291 | + }, |
| 292 | + }, |
| 293 | + { |
| 294 | + alert: 'PostgresHasTooManyDeadTuples', |
| 295 | + annotations: { |
| 296 | + description: '{{ $labels.instance }} has too many dead tuples, which may lead to inefficient query performance. Consider vacuuming the database.', |
| 297 | + summary: 'PostgreSQL has too many dead tuples.', |
| 298 | + }, |
| 299 | + expr: '(sum without(relname) (pg_stat_user_tables_n_dead_tup{asserts_env!="", db_name!~"template.*|^$"}) > 10000) / ((sum without(relname) (pg_stat_user_tables_n_live_tup{asserts_env!="", db_name!~"template.*|^$"}) + sum without(relname)(pg_stat_user_tables_n_dead_tup{asserts_env!="", db_name!~"template.*|^$"})) > 0) >= 0.1 unless on(instance, asserts_env, asserts_site) (pg_replication_is_replica{asserts_env!=""} == 1)', |
| 300 | + 'for': '5m', |
| 301 | + labels: { |
| 302 | + asserts_severity: 'warning', |
| 303 | + asserts_entity_type: 'Service', |
| 304 | + asserts_alert_category: 'failure', |
| 305 | + }, |
| 306 | + }, |
| 307 | + { |
| 308 | + alert: 'PostgresTablesNotVaccumed', |
| 309 | + annotations: { |
| 310 | + description: '{{ $labels.instance }} tables have not been vacuumed recently, which may lead to performance degradation.', |
| 311 | + summary: 'PostgreSQL tables not vacuumed.', |
| 312 | + }, |
| 313 | + expr: 'group without(pod, instance)(timestamp(pg_stat_user_tables_n_dead_tup{asserts_env!=""} > pg_stat_user_tables_n_live_tup{asserts_env!=""} * on(asserts_env, asserts_site, namespace, job, service, instance, server) group_left pg_settings_autovacuum_vacuum_scale_factor{asserts_env!=""} + on(asserts_env, asserts_site, namespace, job, service, instance, server) group_left pg_settings_autovacuum_vacuum_threshold{asserts_env!=""})) < time() - 36000', |
| 314 | + 'for': '30m', |
| 315 | + labels: { |
| 316 | + asserts_severity: 'critical', |
| 317 | + asserts_entity_type: 'Service', |
| 318 | + asserts_alert_category: 'failure', |
| 319 | + }, |
| 320 | + }, |
| 321 | + { |
| 322 | + alert: 'PostgresTableNotAnalyzed', |
| 323 | + annotations: { |
| 324 | + description: '{{ $labels.instance }} table has not been analyzed recently, which might lead to inefficient query planning.', |
| 325 | + summary: 'PostgreSQL table not analyzed.', |
| 326 | + }, |
| 327 | + expr: ' |
| 328 | + group without(pod, instance)( |
| 329 | + timestamp( |
| 330 | + pg_stat_user_tables_n_dead_tup{asserts_env!=""} > |
| 331 | + pg_stat_user_tables_n_live_tup{asserts_env!=""} |
| 332 | + * on(asserts_env, asserts_site, namespace, job, service, instance, server) group_left pg_settings_autovacuum_analyze_scale_factor{asserts_env!=""} |
| 333 | + + on(asserts_env, asserts_site, namespace, job, service, instance, server) group_left pg_settings_autovacuum_analyze_threshold{asserts_env!=""} |
| 334 | + ) |
| 335 | + - |
| 336 | + pg_stat_user_tables_last_autoanalyze{asserts_env!=""} |
| 337 | + > 24 * 60 * 60 |
| 338 | + )', |
| 339 | + labels: { |
| 340 | + asserts_severity: 'warning', |
| 341 | + asserts_entity_type: 'DataSource', |
| 342 | + asserts_alert_category: 'failure', |
| 343 | + }, |
| 344 | + }, |
| 345 | + { |
| 346 | + alert: 'PostgresTooManyCheckpointsRequested', |
| 347 | + annotations: { |
| 348 | + description: '{{ $labels.instance }} is requesting too many checkpoints, which may lead to performance degradation.', |
| 349 | + summary: 'PostgreSQL too many checkpoints requested.', |
| 350 | + }, |
| 351 | + expr:' |
| 352 | + rate(pg_stat_bgwriter_checkpoints_timed_total{asserts_env!=""}[5m]) / |
| 353 | + (rate(pg_stat_bgwriter_checkpoints_timed_total{asserts_env!=""}[5m]) + rate(pg_stat_bgwriter_checkpoints_req_total{asserts_env!=""}[5m])) |
| 354 | + < 0.5', |
| 355 | + 'for': '5m', |
| 356 | + labels: { |
| 357 | + asserts_severity: 'warning', |
| 358 | + asserts_entity_type: 'Service', |
| 359 | + asserts_alert_category: 'failure', |
| 360 | + }, |
| 361 | + }, |
123 | 362 | ],
|
124 | 363 | },
|
125 | 364 | ],
|
|
0 commit comments