-
Problem report
-
Resolution: Fixed
-
Minor
-
4.0.15
-
Sprint 58 (Nov 2019), Sprint 59 (Dec 2019)
-
1
Infrastructure:
429580 Hosts
20084 Host groups
19856 Maintenances
When starting maintenance for all of the hosts, timer gets paralyzed by processing maintenance and locks the configuration cache by paralyzing all zabbix service.
Last message from timer:
# ps ax | grep timer 793 ? R 928:12 /sbin/zabbix_server: timer #1 [updated 120293 hosts, suppressed 244 events in 1888.558713 sec, processing maintenances]
Zero response in strace. Nothing gets written to the log file.
outpud from gdb:
(gdb) bt #0 0x00000000004b3ed0 in zbx_hash_modfnv () #1 0x00000000004b691d in zbx_hashset_search () #2 0x00000000004a937b in dc_maintenance_match_host () #3 0x00000000004aaed9 in zbx_dc_get_host_maintenance_updates () #4 0x0000000000445854 in timer_thread () #5 0x00000000004bc47e in zbx_thread_start () #6 0x0000000000425be3 in MAIN_ZABBIX_ENTRY () #7 0x00000000004b296d in daemon_start () #8 0x0000000000424c63 in main () (gdb)
