I've installed my first LI Agent on linux yesterday to get data from our qmail and pound log to be able to query them in a simpler way.
I saw that on imported log I have various vmw_ fields (vmw_cluster,vmw_datacenter,vmw_host,vmw_object_id,vmw_vcenter,vmw_vcenter_id,vmw_vr_ops_id) that is a nice thing (I hope they'll also be used to integrate alarms into vrops so it's really good to get alerts from an application log directly into the host/cluster/etc).
The problem is that those fields have the wrong value.
My infrastructure is composed by 2 vcenter server ( I monitor both with vrops and LI) and I have SRM configured. The 2 datacenter are replicated so I have VM powered on in DC1 and I see them powered off in DC2.
The logs of the 2 linux server where I have installed the LIAgent for linux report the wrong Datacenter/Cluster/Host/vmid... they report the data of the object in the Naple DC (that is the SRM placeholder machine) Probably because "alphabetically" that vCenter is before the productive ones for those VM (production=ITROM, DR=ITNAP).
It seems a bug on how LI "recongnize" the object (probably it should ignore powered off object with the name of the object reporting the data.... should I hope a ticket about it?
There's any workaround on this?
Very interesting - yes please do.
Yesterday when I did the first configuration I used syslog protocol as it was the only port opened on the firewall from those machine.
Today I've reconfigured them to use cfapi and it seems that they get the correct cluster/hosts/datacenter. Tomorrow I'll do some more tests and I'll open a SR if I'm able to replicate the problem with specific information.
Opened SR 15613721602 for reporting the bug.
did you ever get an answer on this? I'm seeing same issue