Are there templates for zabbix-2.4.X? On your website you only mention 3.X templates
Ad Widget
Collapse
Improved VMWare Monitoring with VmBix
Collapse
X
-
I've sorted the template import issue, but the discovery seems to be failing. I'm seeing a lot of this in the vmbix log.
VmBix.java:2162] No vm named '564d445e-e86f-e73a-1bf4-00d797d5e3fb' found
I'm getting some data returned for the datastores and on the vcentre item itself.Comment
-
Restart the service after changing the value.
Get the vCenter SDK version :
discover the VMs:Code:zabbix_get -s 127.0.0.1 -p 12050 -k "about"
and check a VM :Code:zabbix_get -s 127.0.0.1 -p 12050 -k "vm.discovery[*]"
Code:zabbix_get -s 127.0.0.1 -p 12050 -k "vm.powerstate[VMUUID]"
Comment
-
I think changing the useuuid has fixed it as I'm getting data returned now...
Looking good.
How does it pick up vmtools for centos7 that has open-vm-tools installed? ESX picks it up as foreign, but this says unsupported?
Also can we see the host memory in terms of active and consume etc just as we can for the virtuals?
I assume the user I connect to Vcenter with can be a read only user?
This is already looking much better than the inbuilt one!Comment
-
For the ESX memory, you can add items for the esx.vms.* methods. See the Wiki : https://github.com/dav3860/vmbix/wikiComment
-
Hi,
I'm trying that, but it's returning small values that don't seem to match up with what I can see when i look at the host in vcenter.
i.e. on the ESXi host vmbix[esx.vms.memory,{HOST.HOST},active] is returning 7 MB when I can see that 15GB is active in vcenter.Comment
-
I've removed the scaling, but the values I'm getting don't make sense.
My Host has 128GB of memory and looking in VCentre I see that (in K) I have 102867172 Consumed (the Used reports this as 98GB correctly) but esx.vms.memory consumed returns this as 51% rather than ~77
For active I'm getting 7% returned when there is 13836450K which equates to around 10%
The averages are pretty close to the latest on v center so i know they're not fluctuating...Comment
-
I'd like the absolute values of active and ballooned at least. Consumed, I think, is what is already reported as Used.
The percentages would be ok if I was sure they tied up with what the host itself is reporting. I'm not sure what has gone wrong there.
JasonComment
-
I see. The values you see in the VI client are performance counters. They are described here at the cluster/resource pool/ESX/VM level :
So VmBix is already able to collect these items using the *.counter methods. First, check the available memory perf counters for the ESX hosts :
You will get a list of performance counters. For example, the VI client exposes these items :Code:zabbix_get -s 127.0.0.1 -p 12050 -k "esx.counter.list[ESXUUID]" | grep mem
you can query these values with :Code:mem.granted.average mem.swapused.average mem.sharedcommon.average mem.vmmemctl.average -> for the balloon memory [...]
or :Code:zabbix_get -s 127.0.0.1 -p 12050 -k "esx.counter[ESXUUID,mem.granted.average]"
By default, the values are aggregated with a 300s interval. You can change this in the configuration file with the "interval" parameter.Code:zabbix_get -s 127.0.0.1 -p 12050 -k "esx.counter[ESXUUID,mem.vmmemctl.average]"
Last edited by tatapoum; 04-08-2016, 14:43.Comment
Comment