[ovirt-users] Any monitoring tool provided?

Thu Mar 29 07:26:04 UTC 2018

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hi SHIRLY,

first why zabbix. Zabbix is our primary monitoring tool. Zabbix have
native implementation for VmWare monitoring, which we of course use.
The idea was to use the same monitoring system, alerting for our NOC.

The performance problem is due the how zabbix works, especially zabbix
agent. You could return only one value each query, to to get stat for
VM you have to do  about 10-15 queries. Zabbix agent needs to for the
process ... So overhead is twice, to get the data using api and
forking process.

I see the grafana integration,
https://www.ovirt.org/blog/2018/01/ovirt-report-using-grafana/. This
seems to be dependent on oVirt version, as I see the table names.

So my next steps are:
a)
I took look at the ovirt metrics, looks good. Is it stable on 4.2?
b)
take a look on zabbix sender, but  this have another sets of issues.
For example how to send data olny for some VM/Hosts, not for all if I
do not want to monitor while platform. Data could be collected from
API or DWH
c)
try to do the oVirt poller into zabbix ??

Generally what we need is monitoring and alerting system.
 - alert if Vm is down
 - alert if VM CPU usage is high for period of time
 - alert if there is storage problem /latency, usage, .../

	Peter

On 29/03/2018 08:41, Shirly Radco wrote:
> Hi Peter,
> 
> Using the oVirt API will add load on the engine and may affect the 
> engine performance. We have the DWH that collectd metrics and
> configuration data about the hosts and vms and can be used along
> with Grafana.
> 
> We also added the oVirt metrics store that collects the metrics
> directly from the host about the hosts and vms and from the engine
> about the machine itself and postgres db. This was done in order to
> lower the load on the engine. We also collect the engine.log and
> vdsm.log.
> 
> The metrics are collected to Elasticsearch and can be visualised
> in Kibana and Grafana and are collected in 10 sec interval that
> give you almost real time view. I would love to hear why you prefer
> using Zabbix for collecting the metrics data.
> 
> Best regards,
> 
> --
> 
> SHIRLY RADCO
> 
> BI SeNIOR SOFTWARE ENGINEER
> 
> Red Hat Israel <https://www.redhat.com/>
> 
> <https://red.ht/sig> TRIED. TESTED. TRUSTED.
> <https://redhat.com/trusted>
> 
> 
> On Wed, Mar 28, 2018 at 5:38 PM, Peter Hudec <phudec at cnc.sk 
> <mailto:phudec at cnc.sk>> wrote:
> 
> I have working proof-of-concept. There still could  be bug in 
> templates and not all parameters are monitored, but HOST and VM 
> discovery is working.
> 
> I will try to share it on github. At this moment I have some 
> performance issues, since it's using the zabbix-agent which needs
> to for a new process for each query ;(
> 
> Peter
> 
> On 26/03/2018 08:13, Peter Hudec wrote:
>> Yes, template and python
> 
>> at this moment I understand oVirt API. I needed to write own
>> small SDK, since the oVirt SDK is using SSO for login, that means
>> it do some additional requests for login process. There is no
>> option to use Basic Auth. The SESSION reuse could be useful, but
>> I do no want to add more
> 
>> First  I need to understand some basics from Zabbix Discovery 
>> Rules / Host Prototypes. I would like to have VM as separate
>> hosts in Zabbix, like VMWare does. The other stuff is quite
>> easy.
> 
>> There plugin for nagios/icinga if someone using this monitoring 
>> tool https://github.com/ovido/check_rhev3
> <https://github.com/ovido/check_rhev3>
> 
>> On 26/03/2018 08:02, Alex K wrote:
>>> Hi Peter,
> 
>>> This is interesting. Is it going to be a template with an 
>>> external python script?
> 
>>> Alex
> 
>>> On Mon, Mar 26, 2018, 08:50 Peter Hudec <phudec at cnc.sk
> <mailto:phudec at cnc.sk>
>>> <mailto:phudec at cnc.sk <mailto:phudec at cnc.sk>>> wrote:
> 
>>> Hi Terry,
> 
>>> I started to work on ZABBIX integration based on oVirt API. 
>>> Basically it should be like VmWare integration in ZABBIX with 
>>> full hosts/vms discovery and statistic gathering.
> 
>>> The API provides  for each statistics service for NIC, VM as
>>> wall CPU and MEM utilization.
> 
>>> There is also solution based on reading data from VDSM to 
>>> prometheus
>>> 
> http://rmohr.github.io/virtualization/2016/04/12/monitor-your-ovirt-d
> <http://rmohr.github.io/virtualization/2016/04/12/monitor-your-ovirt-d
>
>
> 
a
> 
>>> 
>>> 
> ta
> 
> 
>> center-with-prometheus
>>> 
> <http://rmohr.github.io/virtualization/2016/04/12/monitor-your-ovirt-
> <http://rmohr.github.io/virtualization/2016/04/12/monitor-your-ovirt->
>
> 
d
> 
>>> 
>>> 
> atacenter-with-prometheus>.
> 
> 
> 
>>> Peter
> 
>>> On 22/03/2018 04:41, Terry hey wrote:
>>>> Dear all,
> 
>>>> Now, we can just read how many storage used, cpu usage on 
>>>> ovirt dashboard. But is there any monitoring tool for 
>>>> monitoring virtual machine time to time? If yes, could you
>>>> guys give me the procedure?
> 
> 
> 
>>>> Regards Terry
> 
> 
>>>> _______________________________________________ Users
>>>> mailing list Users at ovirt.org <mailto:Users at ovirt.org>
> <mailto:Users at ovirt.org <mailto:Users at ovirt.org>>
>>> http://lists.ovirt.org/mailman/listinfo/users
> <http://lists.ovirt.org/mailman/listinfo/users>
> 
> 
> 
>>> _______________________________________________ Users mailing 
>>> list Users at ovirt.org <mailto:Users at ovirt.org>
> <mailto:Users at ovirt.org <mailto:Users at ovirt.org>>
>>> http://lists.ovirt.org/mailman/listinfo/users
> <http://lists.ovirt.org/mailman/listinfo/users>
> 
> 
> 
> 
> _______________________________________________ Users mailing list 
> Users at ovirt.org <mailto:Users at ovirt.org> 
> http://lists.ovirt.org/mailman/listinfo/users 
> <http://lists.ovirt.org/mailman/listinfo/users>
> 
> 

- -- 
*Peter Hudec*
Infraštruktúrny architekt
phudec at cnc.sk <mailto:phudec at cnc.sk>

*CNC, a.s.*
Borská 6, 841 04 Bratislava
Recepcia: +421 2  35 000 100

Mobil:+421 905 997 203
*www.cnc.sk* <http:///www.cnc.sk>

-----BEGIN PGP SIGNATURE-----
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=j0FI
-----END PGP SIGNATURE-----