Performance issue on Zabbix 5.0?

[augustogomez]

We are implementing zabbix to monitor 3 SGI UV3000 systems using their built in snmp agent. Monitoring is basically receive an snmp trap from the SGI, evaluate it and trigger on it if is an actual problem which is basically see if an object on the trap has specific values. Today, while doing some maintenance on the SGIs, we Zabbix started to trigger on traps that where not actual problems. they where just the SGI system telling "I am going up". In other words a flood of traps seem to make zabbix to trigger events.. in fact the events solved immediately.

Here's the actual text from the trap:

11:18:56 2020/07/07 PDU INFO:
transactionid 56
notificationtype TRAP
version 1
errorstatus 0
messageid 0
requestid 336504691
receivedfrom UDP: [192.168.1.120]:43008->[192.168.1.122]:162
errorindex 0
community sgi_uv_traps
VARBINDS:
DISMAN-EVENT-MIB::sysUpTimeInstance type=67 value=Timeticks: (209722735) 24 days, 6:33:47.35
SNMPv2-MIB::snmpTrapOID.0 type=6 value=OID: SNMPv2-SMI::enterprises.59.3.800.100.20.2
SNMPv2-SMI::enterprises.59.3.800.10.10.1.1 type=4 value=STRING: "UV3-0000271"
SNMPv2-SMI::enterprises.59.3.800.10.20.1.1 type=4 value=STRING: "r001i11c"
SNMPv2-SMI::enterprises.59.3.800.30.11.1.1 type=4 value=STRING: "FN8"
SNMPv2-SMI::enterprises.59.3.800.30.11.1.2 type=4 value=STRING: "3696"
SNMPv2-SMI::enterprises.59.3.800.30.11.1.4 type=2 value=INTEGER: 1]

Here's the Item to process the traps:

snmptrap["SNMPv2-SMI::enterprises.59.3.800.100.20.2"]

And here's the trigger, which basically will trigger if the trap indicates a LowerNonRecoverable or UpperNonRecoverable sensor value has been reached. An object vale of 1 indicates the sensor is Ok, and that is the value the actually was triggered (see above):

{SGI-UV3000:snmptrap["SNMPv2-SMI::enterprises.59.3.800.100.20.2"].str(INTEGER: 2)}=1 or {SGI-UV3000:snmptrap["SNMPv2-SMI::enterprises.59.3.800.100.20.2"].str(INTEGER: 7)}=1

Does anyone has experience something similar?