Login to Hyperflex connect failing

[Moon1998]

Hi,

I have recently installed Hyperflex cluster - version 3.5(2b). Everything was working fine, but recently I have issues with login to HX connect management interface. The site loads up, I am prompted for credentials, after providing them it just sits there for quite a time and then I got red message "Check your cluster status or logs for possible solutions".

I am able to log to the cluster mgmt IP with ssh, "stcli cluster info" says cluster is healty:

--

cluster:
allFlash: False
healthState: healthy
capacity: 6.0T

--

It's not a browser issue, I tried from different browsers and computers. It happened before and I was able to fix the issue by rebooting whole cluster. Obviously that's not acceptable solution. Anyone has run into this problem?

Thanks,

Martin

[Clifford Aldan]

Please engage with Cisco TAC to address this issue as they may have to restart some services on the storage controllers to resolve this or it's due to a bug. The TAC engineer will make that determination once they're engaged

[Moon1998 · Accepted solution]

I was able to fix the issue. The whole environment is behind NAT (meaning NAT between HX and my client station I am using for management). Besides NAT for HX connect, I have natted SCVM adresses as well, as this seems to be required for Intersight connection (for the connection HX connect is using IP address of the SCVM, not HX connect cluster IP).

Turns out, this NAT for SCVM caused the issue. When I removed NAT, and left just the one for HX connect, everything works fine. Well, the intersight connection doen'ts work. Plus, I don't really understand why this matters. All key components (SCVM, ESXi, VC) are in the same network, why NAT of SCVM breaks the HX connect? 

Can anyone explain this?

[ShishirDash5204]

Is this issue still persist with hyperflex 4.0(2b).

I am able to take ssh but not able to access through browser.

Any solutions is welcome