Hardware sensors collection failure after V8 upgrade

[Kyrikaze]

Hi,

I just upgdraded from Veeam One 7 to Veeam One 8, and I get in some trouble ;(

I opened a case to the support, but, I would like to resolve my issue this Week-end.
The number of the case is : # 00679362

I have read a similar post but, I tried everything without succes:

- My CIM service are running on my 3 ESXI Host (Vsphere 5.5).
- I checked the status of each Host on Vsphere, and everything is OK
- I can make a Nslookup from my Veeam One Server to each Host and their names are resolved correctly
- I also add their IP and name in the local host file of my Veeam Server to be sure
- I restarted my ESXI
- I restarted my VCenter

Does anyone have some clues ?
I was planning to upgrade my Veeam B & R but I think I'll wait until I resolved this problem first.

Thank you.

[Vitaliy S.]

The reason for this issue should be clearly stated in the log files, so let's not guess (since you've already tried other possible fixes) and wait what our support team can say after reviewing your debug log files.

[raiciks]

Hi,
I have the same problem.
Veeam One monitoring have direct access to ESXi 5.5 hosts! In previous versions all work fine!
Please give real solution, what is the problem?

[Vitaliy S.]

Hi,

Have you tried one of the following solutions?

Thanks!

[raiciks]

We do not use vCenter!
I added ESXi hosts DNS names in Veeam One server host file.
CIM server service is working and ports are enabled on ESXi firewalls. Also veeam can monitor all esxi servers without problem, just doesn't can't collect hardware status. We have HP and DELL servers. This problem are on all servers.

[Vitaliy S.]

If you see all the sensors in the vSphere Client and still cannot collect them via Veeam ONE Monitor Client, please open support ticket with our technical team and let me know your case ID.

[nzorn]

Case # 00679190 here...Seeing this problem after v8 upgrade as well.

I've supplied log files already, had a Veeam Engineer remote into our environment, they also just escalated it to the next tier, and the engineer is going to remote in tomorrow.

[Vitaliy S.]

While you're waiting for the remote WebEx, can you check what is specified in the setting below for the hosts that you cannot retrieve data from?

Code: Select all

vCenter Server config (Configuration -> DNS and Routing -> Host Identification)

[nzorn]

While you're waiting for the remote WebEx, can you check what is specified in the setting below for the hosts that you cannot retrieve data from?

Code: Select all

vCenter Server config (Configuration -> DNS and Routing -> Host Identification)

Yes, the first engineer verified that all of the DNS information was entered correctly. I did add the DNS suffix to my Veeam One server though.

Also I can retrieve data from all of my ESXi hosts, but they randomly trigger an error, and reset/resolve themselves automatically after about 5 minutes.

[raiciks]

If you see all the sensors in the vSphere Client and still cannot collect them via Veeam ONE Monitor Client, please open support ticket with our technical team and let me know your case ID.

Hi, my support case nr. is: 00683521

[Vitaliy S.]

Also I can retrieve data from all of my ESXi hosts, but they randomly trigger an error, and reset/resolve themselves automatically after about 5 minutes.

If you're able to collect hardware data, then DNS configuration should not be the issue. I will ask our dev team to take a look at the latest logs.

Hi, my support case nr. is: 00683521

Thanks for the case ID, will take a look now.

[aleneosp]

Same problem here. Any updates?

[Vitaliy S.]

Alexandre, have you checked for possible solutions mentioned in this thread (2nd page)?

[patrickl78]

I have the same issue on a clean install of Veeam One 8 and one that I upgraded. Both exhibit failure to collect hardware information. I've restarted the CIM services and made sure the account I'm using has access to pull the information. It shows fine in a vSphere client, yet on my Veeam One all i see are errors. Case #00693430

Thanks,
Patrick

[Vitaliy S.]

Thanks for the case, I will follow up it with our support team.

[KimmoJ]

Just upgraded to V8 as well from 7 and have this issue also. I was wondering what was going on since it worked fine in 7. Thought I'd chime in and put myself on the notification list for this thread in case suddenly, a wild solution appears.

[Vitaliy S.]

In version 7 we have used collection via MOB (Managed Object Browser) and CIM-SMASH protocol. In version 8 we have switched to CIM-SMASH protocol which VMware says a preferred way of monitoring hardware sensors. If CIM service is misbehaving for some reason, you will see this kind of errors.

Currently, we are exploring other methods of collecting hosts hardware.

[rferrigno]

I'm having the same problem. I've been working with support for a few weeks now with no resolution.

[Vitaliy S.]

Hi Robert, can you please tell me your case ID?

[rferrigno]

After reading more in depth information on how the CIM server works I was able to solve this in my environment maybe it will be helpful to some of you. In order to gather hardware data Veeam uses CIM and makes an HTTPS call on port 5989. So I went into IE and typed this in https://IPofesxiserver:5989 and I got a certificate error followed by a 501 error which is what you should get. Next I tried via the FQDN which Veeam uses like this. https://esxihost.company.com:5989 and it failed outright. No cert error. I was able to ping the FQDN and use nslookup so I decided to look at the settings in IE. It ended up being the proxy settings. Internet options-connections-Lan settings. I removed the proxy and everything started working. You could also add the FQDN to the bypass if you need to keep the proxy enabled. That's it problem solved.

[Vitaliy S.]

Hi Robert, thank you for sharing your knowledge, I will forward this info to our dev team.

[stingray75]

We use two DNS systems at my company, one is AD integrated (amer.intra zone) and the other is Infoblox (us.intra zone) which is used for all hosts (Linux, Windows, AIX etc). Our ESX hosts are only registered in Infoblox us.intra as they are not Windows hosts - this is global policy, and we're not allowed to create static entries in AD DNS. Workstations/servers have amer.intra and us.intra in the suffix search list so names in both zones are always resolve.

In order to use AD integration, we specify Configuration -> DNS and Routing -> Domain as amer.intra. However, the problem we have now is that Veeam tries to resolve servername.amer.intra and this fails. Could there be an option to ignore the domain setting so it resolves servername to servername.us.intra, or force it to add .us.intra instead?

I tried using hosts files and it works, but we have nearly 100 hosts and the environment can be fluid, so I really don't want to use hosts files.

Thanks

Matt

[Vitaliy S.]

Hi Matthew,

What if you specify DNS server (Infoblox) as a secondary server on your Veeam ONE server in the Window Server settings? Based on my knowledge it should also solve the DNS resolution problem.

Thanks!

[init-s]

Hi,

I can update this thread with my case, for the same problem: case ID 00677996 for VEEAM and case ID 1935005 for OVH (the infrastructure is managed by OVH that has acknowledge the problem).

[Vitaliy S.]

Guys, I just wanted to let you know, that our support team now has a debug tool (kudos to our dev team for that) that verifies CIM service availability and has an option to switch to another way of collecting host hardware sensors. We will validate this tool (need to make sure that it works in every environment), and then will roll-out a public update for this issue. Thanks everyone who has submitted their case IDs.

[marcode]

Same problem here, support case is 00702529

[ihopley]

Same problem here: Case 00702943.

[Vitaliy S.]

For everyone who is experiencing the same issue, ask your engineer to validate sensors status with a debug tool. This will help us to nail down CIM service health state and response times.

[tlawler]

I'm experiencing the same issue. My case #00700666

I have confirmed the servers ability to ping the domain & names in DNS setting on VCenter Server.

[Vitaliy S.]

What about our support utility that was built for verifying sensors availability ? Does it collect the data?