VMWare event collected


I connected MIQ to 2 vCenters with about 5000 VM each. Events and CU were being collected.

The vCenters were overloaded up to DoS with the collect performed by MIQ.
I plan to disable CU collection to reduce the load on the vCenters and MIQ.

But i need the events to be collected to have new VM discovered.

My question is about which events are being collected by MIQ with the vCenter provider?


Yes with that many VMs you’ll need to plan and scale out your ManageIQ deployment carefully. I’d recommend a separate zone per VMware provider, and for resilience you should be planning at least 2 MIQ appliances per zone. This’ll help spread the worker load between appliances.

Disabling C&U will certainly help reduce the load on the vCenters and MIQ appliances as these are probably the most resource-intensive workers. I’d have thought that if you do this and scale out your MIQ appliances as I suggested, the load on the vCenters from collecting events should be minimal.

There’s a Red Hat CloudForms reference architecture called Deploying CloudForms at Scale that should be also applicable to ManageIQ, even though it’s a few versions old now. It covers zones and scaling.

Hope this helps,

Thank you for your response pemcg.
This is much appreciated.

I was wondering whether I could also tweak MIQ to collect only some events needed for VM inventory, instead of all the vCenter events. It looks like the code collects all the events.


It is possible to blacklist some provider events that you don’t want processed, but I don’t believe that it’s possible to explicitly list events that should only be collected.

If you’re in pgsql, try:

vmdb_production=# select event_name,provider_model from blacklisted_events;


This looks very good to reduce the events collect workload:

vmdb_production=# select event_name,provider_model from blacklisted_events where provider_model=‘ManageIQ::Providers::Vmware::InfraManager’;
event_name | provider_model
AlarmActionTriggeredEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmCreatedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmEmailCompletedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmEmailFailedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmReconfiguredEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmRemovedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmScriptCompleteEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmScriptFailedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmSnmpCompletedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmSnmpFailedEvent | ManageIQ::Providers::Vmware::InfraManager
AlarmStatusChangedEvent | ManageIQ::Providers::Vmware::InfraManager
AlreadyAuthenticatedSessionEvent | ManageIQ::Providers::Vmware::InfraManager
EventEx | ManageIQ::Providers::Vmware::InfraManager
UserLoginSessionEvent | ManageIQ::Providers::Vmware::InfraManager
UserLogoutSessionEvent | ManageIQ::Providers::Vmware::InfraManager
(15 rows)


So adding new entries is a solution to disable collection of unused events.
Need to know which events are really needed.


The analysis profile are set in settings =>MIQ regions
Select “host default” => “configuration”, then you can add or delete entries to be analysed from what you need.
You can also create a profile for Vms from “sample”
Hope this help

Hello Jpj,

What you mentioned is related to Host profile.
I was requesting how to reduce the event types collected by the VMware provider.