used by the system (in kilobytes). • User macro resolving has been moved from the preprocessing manager to preprocessing workers. Agent 2 configuration Buffer size The default value of the BufferSize between service checks. A new discovery manager process has been added along with a configurable number of discovery workers (or threads). 15 The discovery manager processes discovery rules and creates valuecache - value cache statistics preprocessing - preprocessing manager statistics alerting - alert manager statistics lld - LLD manager statistics locks - list of mutexes (is empty on BSD systems)

templates are now available for out-of-the-box monitoring: Cisco • Cisco UCS Manager SNMP - monitoring of Cisco UCS Manager via SNMP. DELL PowerEdge • DELL PowerEdge R720 by HTTP - monitoring of DELL valuecache - value cache statistics preprocessing - preprocessing manager statistics alerting - alert manager statistics lld - LLD manager statistics locks (since 5.0.5) - list of mutexes snmp_cache_reload The scripts are located under directory misc/init.d. Server process types • alert manager - alert queue manager • alert syncer - alert DB writer • alerter - process for sending notifications • configuration

in case of Zabbix agent, SNMP, JMX items. Pollers, unreachable pollers, JMX pollers and the IPMI manager would update host availability directly in the database with a separate transaction for each host new availability manager process has been introduced. All processes queue host availability updates to the availability manager and that queue is flushed by the availability manager to the database every templates are now available for out-of-the-box monitoring: Cisco • Cisco UCS Manager SNMP - monitoring of Cisco UCS Manager via SNMP. HP ProLiant • HPE ProLiant BL460 SNMP - monitoring of HPE ProLiant

external systems In this development, two new processes have been added to Zabbix server: connector manager and connector worker. A new Zabbix internal item zabbix[connector_queue] allows to monitor the count valuecache - value cache statistics preprocessing - preprocessing manager statistics alerting - alert manager statistics lld - LLD manager statistics locks - list of mutexes (is empty on BSD systems) all proxies secrets_reload Reload secrets from Vault. service_cache_reloadReload the service manager cache. snmp_cache_reloadReload SNMP cache, clear the SNMP properties (engine time, engine boots

for Zabbix server performance, as data from proxies will no longer go through the preprocessing manager, the affected proxies will require more resources than before. 10 Warning: The change affects consisting of a config- urable number of low-level discovery workers and a low-level discovery manager. This change greatly reduces low-level discovery impact on data gathering (poller, trapper processes) start-up/shutdown. The scripts are located under directory misc/init.d. Server process types • alert manager - manager of alerter tasks • alerter - process for sending notifications • configuration syncer - process

HTTP • Kubernetes cluster state by HTTP • Kubernetes API server by HTTP • Kubernetes Controller manager by HTTP • Kubernetes Scheduler by HTTP • Kubernetes kubelet by HTTP To enable Kubernetes monitoring by HTTP • Azure by HTTP • Control-M server by HTTP • Control-M enterprise manager by HTTP • Veeam Backup Enterprise Manager by HTTP • Veeam Backup and Replication by HTTP See setup instructions for valuecache - value cache statistics preprocessing - preprocessing manager statistics alerting - alert manager statistics lld - LLD manager statistics locks - list of mutexes (is empty on **BSD* systems)

be sent. In the new version, parallel processing of alerts is implemented. There is a new alert manager process that can distribute alerts to several ”worker” processes, if needed. Media types are processed Retry interval: • Functionality of the database watchdog process has been merged into the alert manager and the watchdog process itself removed. 6 Improved notifications 6.1 Notification on problem acknowledgement number of IPMI pollers this could easily overloaded BMC controllers. Since Zabbix 3.4, a new IPMI manager process has been added to schedule IPMI checks by IPMI pollers. Now a host is always polled by the

valuecache - value cache statistics preprocessing - preprocessing manager statistics alerting - alert manager statistics lld - LLD manager statistics locks - list of mutexes secrets_reload Reload secrets The scripts are located under directory misc/init.d. Server process types • alert manager - alert queue manager • alert syncer - alert DB writer • alerter - process for sending notifications • configuration poller for icmpping checks • ipmi manager - IPMI poller manager • ipmi poller - poller for IPMI checks • java poller - poller for Java checks • lld manager - manager process of low-level discovery tasks

templates New templates are available now: • Control-M enterprise manager by HTTP • Control-M server by HTTP • Veeam Backup Enterprise Manager by HTTP • Veeam Backup and Replication by HTTP 16 See setup valuecache - value cache statistics preprocessing - preprocessing manager statistics alerting - alert manager statistics lld - LLD manager statistics locks - list of mutexes (is empty on **BSD* systems) all proxies secrets_reload Reload secrets from Vault. service_cache_reloadReload the service manager cache. snmp_cache_reloadReload SNMP cache, clear the SNMP properties (engine time, engine boots

start-up/shutdown. The scripts are located under directory misc/init.d. Server process types • alert manager - manager of alerter tasks • alerter - process for sending notifications • configuration syncer - process ipmi manager - IPMI poller manager • ipmi poller - poller for IPMI checks • java poller - poller for Java checks • poller - normal poller for passive checks • preprocessing manager - manager of preprocessing process for collecting internal server statistics • snmp trapper - trapper for SNMP traps • task manager - process for remote execution of tasks requested by other components (e.g. close problem, acknowledge