Cache info/summary idsFix tableDeduplicate info/summary from alert transitionsAdd job_running flag and cleanup handling to health event loopQueue store sql statements in metadata threadRemove unnecessary health check condition from event loopAdd new job names for alert host snapshot and process eventsPrepare ACLK OP to handle alert processing per hostRemove transactions for nowFix compile errorRevert serialization testAdd opcodes for Health Pause/ResumeSerialize alert transition processing / wrap in transaction (test)Set fixed timer intervals for health event loopAccurate scheduling health run after maintenance cannot be done yetRemove unused thread-local variable for health thread and simplify SQL statement preparationRefactor job submission logic to improve worker handlingSimplify job execution handlingMaintenance task should try to keep the health run scheduleImplement worker data pool to avoid memory allocationsFix delay calculation in health_host_register functionFix wrong memory allocationsRefactor job management to use dynamic worker data allocationCleanup workersRefactor health_host_initialize to remove delay parameterRemove VACUUM command from database health cleanupIf cleanup is running returnRefactor health job schedulingPrevent scheduling job if we reached the limitConfigure init, maint and cleanup health thread count based on the configured health threadsSchedule a host health init when we one is completedFix maintenance reschedulingAdd worker jobsCode cleanupAdd health thread configRun maintenanceRebase and code cleanupTransient alert entries are queued for deletion at the end of the evaluation loop (making sure they are processed and saved)Host unregister with cleanupNew job for host rrdcalc cleanup during unregisterstream receiver will schedule a host unregister and rrdcalc cleanup nowConfigure max threads per job (needs improved config)Reschedule alert evaluation if rrdcalc cleanup is pendingCollect all alert transitions and save in a batchRemove alert transition store from metadata event loopAdd host heath maintenance job (not used yet)cleanup_health_log removed from metadata threadMigrate health tables and aclk tables to new databases. Drop old tables.Use attach_databaseStore alert transitions to the new databases (includes queues and version book keeping for the cloud)Add sql_init_databases to configure all databases during startupDo not create aclk and health tables in netdata-meta.dbNew netdata-health database for alert transitionsNew netdata-aclk database to store transitions that need to go to the cloudFix compilation with internal checksRemove static threadInit and shutdown healthRegister/Unregister hosts for healthHandle resume from suspension / health delay of child reconnectRemove static health threadNew health event loop