Jun 15, 2020

I am having a recurring problem where by the number of VM's reported back after a test is not consistent. I am using the default easy run option (have also found it with other custom runs). It deploys 40 VM's for this test but often only reports around 35 in the results. It's the same with both FIO and VCBench. I also checked that I can ping all 40 VM's from the HCIBench VM during the test run. The logs state is faults to SSH to the missings VM's. I am trying to compare the impact of encryption on an 8 node stretched cluster. Any advice greatly received! :)

Jun 19, 2020

this looks like network issue, or some host was disconnected when the testing was running? Could you help to repro this issue and send all hcibench logs to vsanperformance@vmware.com?

Jun 22, 2020

Thanks, Logs sent! No host network issues or disconnected hosts seen during the tests.

May 28, 2020

Trying to run a load on 480 Datastore with the latest version of HCIBench. The /opt/automation/lib/pre-validation.rb script gets to exactly 62 Datastores and last message displayed in the logs are

2020-05-28 22:56:10 -0700: Datastore vodafoneDS3902 type is VMFS
2020-05-28 22:56:10 -0700: Checking If Datastore customDS3902 is accessible from all the hosts of Cluster01

I have confirmed that all Datastore are accessible from all hosts, but the validation does not go past 62 Datastores.

Even if I choose other Datastores Im not able to get past 62. Is this the limit?

Jul 14, 2020

I think I was able to locate the root cause, this will be fixed in the upcoming release, please stay tuned.

May 29, 2020

How much memory is used by the HCIBench appliance when you try to run the test?

May 27, 2020

While using HCI BENCH, i was able to run one easy run report, but since then, i have been trying to utilize the easy run for the sake of argument and i am able to save the config but not able to get the test to validate. Usually there is an long, long delay in validating the config and the test never starts. I try to just and start the test and nothing seems to ever start, but errors out. i received an error about inter cluster connectivity but have not been able to reproduce that error. (using vxrail hardware for this)

May 28, 2020

here is the message that i am getting
no matches for "hosts/*/vms/hci-tvm-*"
2020-05-28 08:30:14 -0700: IP Assignment failed or IP not reachable
2020-05-28 08:30:14 -0700: IP Assignment failed or IP not reachable
2020-05-28 08:30:15 -0700: IP Assignment failed or IP not reachable
2020-05-28 08:30:15 -0700: IP Assignment failed or IP not reachable
2020-05-28 08:30:15 -0700: IP Assignment failed or IP not reachable
2020-05-28 08:30:15 -0700: IP Assignment failed or IP not reachable
2020-05-28 08:30:15 -0700:
2020-05-28 08:30:16 -0700: no matches for "hosts/*/vms/hci-tvm-*"
2020-05-28 08:30:16 -0700: ------------------------------------------------------------------------------
2020-05-28 08:30:16 -0700: Cluster inter-connectivity validation failed, please check the hcitvm logs in /opt/automation/logs/prevalidation or <a href="https://*.(*.*.*8443/output/hcibench_logs/prevalidation/" target='_blank'>here</a> to identify the issue
2020-05-28 08:30:16 -0700: ------------------------------------------------------------------------------

May 29, 2020

This looks like a connectivity problem. Are the test VM being created, and are they booting into an OS, or are they stuck at the PXE boot?

May 11, 2020

Hi there, I have a unique situation in that we have a an Cisco ACI fabric across two locations. This has forced us to have seperate Distributed Virtual switches for each side of the stretched vsan cluster. They are the same layer 2 networks but they are on separate distributed switches that are deployed by ACI. The port groups are named the same within each site. We get the error " Network has duplication". Is there a way to specify the networks on each site or possibly ignore these pre-req check? Thanks in advance!

May 11, 2020

If you skip the check and just run the test, does it execute, or do you have problems with the worker network configuration?

May 28, 2020

it seems as if i can delete the ovf and redeploy the ovf and it will one run time. if i try and run subsquent times, they all get the error message above

May 12, 2020

Hi Charles,

If we just go ahead and run the test it creates a few of the worker vm's but fails on a bunch of them as well. The deployment log on the working host states "The provided network mapping between OVF networks and the system network is not supported by any host."

May 12, 2020