VMs and installation too much slow

[vm_learner07]

Hi there,

I recently started a new job in a company and this one has a VM structure that includes some hosts ESXi 6.0. One of these ESXi hosts is too much slow (really, really slow). Whenever I try to install a VM, this installation takes a long time to finish. For example: I've just tried to install a VM with Windows Server 2012. It started to install about one hour and a half time ago, and so far it didn't finish. Another example: A few days ago, I tried to deploy vCenter Server on this host, but it didn't work properly. The message was "A problem occurred during RPM installation: Timeout". Detail: This process took all day long. Others hosts don't have this problem.

What can I do to fix this problem?

Regards.

[daphnissov]

Two things you must have when deploying vCSA:

1. Forward and reverse DNS records and resolution.
2. NTP timekeeping between the workstation from which you are running the deployment; the target ESXi host; and the vCSA itself during installation.

Without these two things the deployment will not succeed.

[vm_learner07]

daphnissov​

Thanks for answering. So, I just checked these configurations, and everything is Ok. Actually, deploying VCSA is one the issues that it is happening. Anything that I do is very, very, very slow. When I log on a Windows Server VM for example, when I click an icon, it takes a long before opening the program...all of vms in this host show up this behavior...it's weird...

Regards.

[daphnissov]

Then on top of those things you probably have poorly performing storage.

[vm_learner07]

daphnissov​,

I must agree with you. Well, how can I check my storage performance? Is there a way, a tool, whatever, that I can use?

Thanks in advance.

[daphnissov]

Performance charts, graphs, esxtop, etc

[vm_learner07]

daphnissov​,

Thank you. I'm checking datastores. But, I was wondering: why is it happening only on one host, meanwhile others hosts that using same datastores are ok?

Regards.

[daphnissov]

There are too many variables to know why. Some possibilities include incorrect multi-pathing on this host; suboptimal advanced settings, resource contention on this host, hardware/firmware issues depending on what the storage is, etc.

[vm_learner07]

daphnissov​,

I understand. I'm gonna figure it out. I'm trying troubleshooting that with tools that you mentioned. Thank you.

[Dave_the_Wave]

You never mentioned your hardware setup, but I bet the problem is there.

ESXi really hates raid cards that are missing batteries.

[vm_learner07]

Hi Dave_the_Wave​,

Thank you for helping me.

     I'm sorry, I really didn't mention it. I have two hosts ESXi: Both are a Dell PowerEdge R720, 96GB memory, and the DataStores are attached in a VNX 5300 Storage. I have 4 DataStores, each one with 1000 GB space. The connectivity is via Fibre Channel.

     So, in the last days, I've been performing some tests. I deployed a VM directly on the host (on the host where the problem occur). This deploy worked well, I didn't have any problem with this. Then, I tried to install the same VM, same configuration, but I did it in one of the datastores attached on the host (obs.: All of the datastores are attached on both hosts), and...surprise: this VM ran too much slowly. I am thinking this problem is occurring the communication between Host and Storage...but I don't know which problem it is...I'm trying discovering it...however, I don't have too much experience with physical adapters and storage...

Well, the prints below are about the vmkernel.log e vmkwarning.log. The host 224 is the host where the problem is occurring. The host 223 is ok. I think something wrong is happening with vmhba66 (in a comparison between Host223 e Host224) on the host 224:

Host 223 (Host ok):

Host 224 (Host with issues):

Best Regards.