Refresh Hardware Information

[wbn]

NWORKS IS CAUSING THE HOST TO "Refresh Hardware Information" TO MANY TIMES. SEEMS THIS IS CONSTANTLY POPPING UP ON THE VIRTUAL HOST. WHAT CAN I DO TO DECREASE THIS?

Wayne Braden
U.S. Department Of Energy

[Alec King]

Hi Wayne,

This event "Refresh Hardware Information" is required due to a known limitation in the VMware API....the vCenter API does not automatically keep the current hardware status of the ESX hosts.

So in order to have current status of power supply, temperature sensors etc, we have to call an explicit "Refresh Hardware" function for each Host; and this logs an event, as you can see.

We do split the refresh calls over 6 intervals, by default. You can configure the interval 'spread'.The setting is not available in our UI, but you can create this registry key on the nworks Collector(s) -
HKEY_LOCAL_MACHINE\SOFTWARE\nworks\VIC\Collector\GlobalSettings
(for 64-bit system, HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\nworks\VIC\Collector\GlobalSettings)
Create DWORD setting called 'IntervalCount' and set this to your required interval spread.

For example, if you had 24 ESX Hosts, and you set IntervalCount to 24, then one Host would be refreshed every interval (5 mins by default)
If you set IntervalCount to 12, then 2 Hosts would be refreshed every interval, and so on.

You can turn off the refresh hardware event completely - however I would not recommend that, as you would have no visibility of the hardware status of ESX servers.

I'd also recommend you don't set the interval too large; as you will experience a delay in getting hardware alerts into SCOM.

Note the above change only affects Hardware Status alerts. Other aspects of VMware data (vCenter events, Host & VM performance metrics etc) is not affected.

There are more details in the Release Notes for nworks MP 5.0.3, found here http://www.veeam.com/files/release_note ... _5.0.3.pdf

Hope the above answers your question - please let me know,

Regards

[IanP]

Hi,

I appreciate we require this hardware refresh,
BUT
1. It fills the logs - thus removing other vital information from the VC view pane!

why can't it be reported in the negative i.e. It can't refresh the hardware?

Then we could fix the error.

But reporting every success......is this not over kill?

thanks

Ian

[Alec King]

Hi Ian,

I do understand your frustration with the Refresh Hardware events - but this really is a VMWARE limitation.

- If we do not call the refresh function via the VI-API, then the hardware data is not current, and you would miss alerts on power supply, memory error etc.
- if we DO call the refresh function - then VCENTER logs an event. It is not nworks that logs it - we have no choice in this.

However - I am happy to report, that in nworks 5.7 we have a completely new method of gathering hardware data. We have implemented CIM-XML protocol, so we can now talk direct to ESX hosts to get hardware data, and there are no more Refresh events in the VI Client!

Our 5.7 version is shipping very soon - only days away now. Please take a look at the What's New document, available here - http://go.veeam.com/veeam-nworks-v5.7.html. As well as new hardware functionality, there is lots of good stuff in 5.7 - coming very soon!

Cheers,
Alec

[sepj12927]

Hi,

We are experiencing a similar issue, we're getting a lot of "Refresh hardware information" events on one host out of ten VMWare hosts, in the same cluster, being monitored via the same collector.

How come this is happening? I can't recall that we've changed configuration related to that host.

We are running MP 5.7 so we shouldn't even see those events if I understand correctly.

/Per

[chrisdearden]

Might be worth checking the local firewall settings on that host and ensure that port 5989 is open between the host and the collector. I've observed cases where it was inadvertently blocked by a host profile.

[sepj12927]

I've checked and manged to connect to the CIM port from my collector but I still have the "refresh hardware" events. Any suggestion on how to troubleshoot this?

[sergey.g]

Hello Per,

Could you tell us which version of vCenter/ESX you are using? In case of 5.1 some additional procedures are required.

Also it could be a permissions issue, check that account you are using in nworks have all necessary permissions in vCenter. Check Installation Guide for additional information.

http://www.veeam.com/nworks_mp_for_mom_ ... ide_pg.pdf

Thanks,

[sepj12927]

The account I'm using do have the required permissions, it was setup with suprtsiosion of a Veeam employee.

[sergey.g]

Hello,

I'm afraid we need more information to understand what exactly is happening in your enviroment. It would be great if you can open a support ticket and send us logs from your nworks UI.

If you can connect from the collector server to the host and you have permissions configured correctly, the only reason for the issue would be 5.1 build of ESXi. If you do not have 5.1 hosts, then it's something weird and we need more information at least logs from all collector servers.

Thanks.

[albertwt]

Hi,

Is there any update in how to turn off this issue, becaue it is running using DOMAIN\Administrator account?



I'm using VeeamONE version 9.0 Update 1 on Vcenter 5.5U3 with ESXi 5.1 U3 so not sure why is this still happening quite reqularly.

I've checked in the Veeam Enterprise Manager Service website/console there is nothing is configured running using the DOMAIN\Administrator account.

[Alec King]

Hi,

You have posted this question in the Veeam MP forum - not the Veeam ONE forum...however, I do have the answer anyway

You need to right-click on the vCenter Server object in Veeam ONE Monitor Client and then you can change the connection account.

[albertwt]

Hi,

You have posted this question in the Veeam MP forum - not the Veeam ONE forum...however, I do have the answer anyway

You need to right-click on the vCenter Server object in Veeam ONE Monitor Client and then you can change the connection account.

yes Alec,

That does the trick.

Thank you for resolving this issue

[Alec King]

You're welcome!