[V13] WebInterface and Console not working

[stevenrodenburg1]

Dear Sirs,

I downloaded and deployed the V13 appliance (VeeamJeOS_13.0.0.4967_20250822.iso) in the lab today without any problems. Networking with a fixed IP, NTP timesync, OTP via MS Authenticator, everything works. I can login to port 10443 and optically, everything works. The appliance has Internet access (update checking works for example).
ISO Checksum is the same as stated on the download-page.

WebUI is enabled (as it is by default).
The hostname I entered during Install is the FQDN: veeambr13.mydomain.local (the mydomain.local part is made up to anonymise it)
DNS A and PTR Records exist and are correct.

When I try to access the WebUI, assuming it's simply https://veeambr13.mydomain.local , I get a webpage saying:

"HTTP Server Test Page
This page is used to test the proper operation of an HTTP server after it has been installed on a Rocky Linux system. If you can read this page, it means that the software is working correctly."


So maybe there should be something behind https://veeambr13.mydomain.local like /admin or a port, other than 10443 or 443. I can't find any documentation on this, neither in the docs, KB's or in a blog or youtube video somewhere. So I therefore assume it's just plain port 443 and nothing after the hostname like "/admin/ or anything like that.
So far, I have not gotten the normal webUI to work (keep getting that http server test page).

I then installed the console on a Windows 10 22H2 Desktop. Started it, connect to https://veeambr13.mydomain.local which prompts me to accept the thumbnail, which I do.
When I click on "Connect", I always get the error "Failed to connect to the backup server: Failed to connect to identity service (https://veeambr13/) - NotFound". I never get to the point where I can enter a username etc.

I'm at a loss. I re-installed it 3 times over, each time with the exact same results.

[Gostev]

The Web UI address is displayed right in the software appliance console, as can be seen on this screenshot.

If you can't access it using these addresses, then it can only be due to DNS or networking in your environment, as this is not a known issue and it has never been reported by anyone else after many thousands downloads and installations.

Also, I'd like to kindly ask that you follow the first paragraph of forum rules with your future contributions to this community. We will be enforcing this rule more strictly in the next few weeks to avoid noise of environment-specific issues and ensure forum members have a chance to see useful discussions about real problems and actual product bugs (if any) instead of facing a swarm of topics about irrelevant environment-specific issues, which has been the case for virtually every "issue" posted here in the past 10 days.

[stevenrodenburg1]

Hi Anton,

Yes I see those. Port 10443 works normally. When I enter https://veeambr13.mydomain.local:443 (or https://veeambr13:443 or https://<the-ip-address:443 ) I always get that webpage with:

"HTTP Server Test Page
This page is used to test the proper operation of an HTTP server after it has been installed on a Rocky Linux system. If you can read this page, it means that the software is working correctly."

One can normally leave the :443 out by the way as browsers assume that HTTPS:// is referring to port TCP 443. But I added it anyway, to no avail.
I swear to all that's holy. After 3 Installs, this is what I keep ended up with.

Same goes for the console and that "Failed to connect to the backup server: Failed to connect to identity service (https://veeambr13/) - NotFound" error.

About your remarks concerning forum rules. I know the rules. Been here for a looong time. I'm the humble opinion that my topic is a valid question and because I cannot find anything related on the web so far, thought it would be helpful to post it here. I apologise if i'm wrong.

[Gostev]

It does seem like some basic DNS or network connectivity issue. Please open a support case if you need assistance with troubleshooting. This is not a support forum for environment-specific issues nor it is effective to try and troubleshoot networking problems over forum posts.

[stevenrodenburg1]

DNS is fine. Tested name-resolution from inside the appliance as well. /etc/resolv.conf , nameservers, searchdomains all good. Can resolve everything nicely from the CLI (enabled SSH to be able to test).
Firewalling etc. not relevant as in the lab, it's all in the same subnet.

Thanks for your time. I'll open a support-request.

[Gostev]

Ah, those famous first words "DNS is fine" we'll see about that!
Please share the support case ID when you get a chance.

[stevenrodenburg1]

After 35 years in the Business, the dreaded "DNS is fine" statement is very well known to me Anton

I asked our Partner Manager how to get support as we run with Partner/Reseller Lab licenses (or whatever they are called) and trail-licenses and for those, the "create support case" webpage says "Nope".
Or I drive to Baar (which is a short drive) with a case of beer and bribe myself into a support case LOL

[Gostev]

Weird, trial licenses most definitely provide ability to open support cases, may be check to make sure they are actual and not expired long ago.

[stevenrodenburg1]

Little update: I have to wait for our partner manager to become available (traveling).
In the meantime, I might have found a bug (I can reproduce it): when installing, I change the Network Interface from DHCP to Fixed and under "Advanced networking", I enter the two DNS Servers and the Search-domain, then let the installer continue.

Afterwards, when logged into the port 10443 WebGUI and editing the interface, the DNS Server field is blank (there is no Search-domains field by the way). I edit, enter 1 DNS server (don't know how to do multiple), save it, return, still blank. Enter, save, return, blank again. This webgui does not save the DNS Server. Can repeat this over and over. Always ends up blank.

Then, when I go to the native text-based console, edit the advanced network-config, the search-domain I entered is still there, but the DNS fields are gone (not configured). In other words, the installer removed the contents from the DNS Server fields but left the Searchdomain-field intact.
I then add the two DNS servers (just like I did during install), save, go back, still there. Reboot, go back in, still there. So now, for the very first time, the appliance has valid DNS servers for itself.

But the problems "http test page" instead of the normal port 443 webgui and the Console error "cannot connect to identity server, not found" are still there.

I will now delete the appliance VM and re-deploy and keep the Network-settings at the DHCP defaults and just go next-next-finish. See what happens.

[Gostev]

So, famous ending?
"It's always DNS"

[stevenrodenburg1]

"So, famous ending?
It's always DNS"

Yeah but in this case, the appliance-installer makes a bubu with it's DNS config don't you agree.
When I the interface to "static" and type in the DNS Servers and after finishing, they got axed by the installer, that's not me doing it wrong.

So I tried Installing with DHCP. Afterwards, in the 10443 WebGUI, all Networking Interface stuff now says "Automatic" (as expected).
Result: Exact same two problems. No difference at all.

Seriously mate, I would not know how to "install it wrong" even on purpose. I'm a good boy, I follow the instructions, the DNS Server in the Lab has the correct A and PTR records, resolves correctly in every direction so that cannot be it.

It's a flat L2 network, The WebGUI on port 10443 works normally but 443 comes with this "Rocky Linux HTTP Test page" again and the Windows Console bombs out with that "cannot connect to identity server, not found" error, again after accepting the SSL Thumbnail (which proves it talked to the appliance).

I have done 9 deployments now. I'll have to wait until Thursday for our partner manager to return so he can arrange a case and I can show it to an engineer. Again, how can one install this thing incorrectly, it's so basic (which is a good thing). You can't screw it up. It's an appliance. No strange Windows (security-related) GPO's to mess things up like on Windows Server and sending you on a wild goose chase.

Telling me how DNS or Ubuntu or RedHat Linux (incl. CentOS, Alma, Rocky) works is like telling a 5th generation farmer how to milk cows. Redhat v2 was my first kiss. Yet I cannot get this Appliance to work properly and there is nothing I can influence during installation to "break it" as long as the Host and IP-Information is all verifiably correct.

Don't wanna be Mr. negative and grumpy. I'll wait for an engineer and then let's see. Maybe I did manage to screw it up 9 times in a row.

[dali@iae.nl]

The DNS settings not saving behaviour I also added in: veeam-backup-replication-f2/updates-the ... 00052.html
Veeam says (by Dima) they will keep an eye on this behaviour.

[stevenrodenburg1]

Dali wrote in his topic: "From the logs Veeam found that the Veeam Updater is repeatedly failing to connect to the VBR Identity Service at 'vbr03:443'."

I get roughly the same message when a v13 Windows Console on a PC tries to connect to the VBR Appliance.

[BrianBuchanan]

I downloaded and deployed the V13 appliance (VeeamJeOS_13.0.0.4967_20250822.iso) in the lab today without any problems. Networking with a fixed IP, NTP timesync, OTP via MS Authenticator, everything works. I can login to port 10443 and optically, everything works. The appliance has Internet access (update checking works for example).
ISO Checksum is the same as stated on the download-page.

Just thought I'd mention that there are two different ISOs. VeeamJeOS_13.0.0.4967_20250822.iso is the Infrastructure and Hardened repository ISO and there is no web console. (other than /updater) The other ISO is VeeamSoftwareAppliance_13.0.0.4967_20250822.iso and it is the Backup and Replication and Enterprise Manager ISO. It has a a web console.

[Gostev]

Actually, both VSA and VIA ISOs are nearly the same (except added software) and they both do have the Host Management web console. However, depending on the VIA deployment option the web console may be disabled by default for added hardening. And even if you enable it using TUI to perform some management, it will disable itself after some time.

[stevenrodenburg1]

Omg. So apparently I was given the "Veeam Infrastructure Appliance" 13.0.0.4967. Can't blame anyone, the name sounds perfectly logical. The name does raise any red flags "hmmm don't think that's the right one bro".

But it was the wrong ISO the entire time. There is another. I needed the "Veeam Software Appliance" 13.0.0.4967. Seriously???
Well, i'm downloading "the other one" now. Gonna deploy it and it will probably "fix all my problems". Can I get those many frustrating hours of my life back pls...

Seriously, somebody fix the naming of these ISO's, make it very clear what is what and what they are for...

[stevenrodenburg1]

Update: the "Veeam Software Appliance" 13.0.0.4967 works.
It does not have the "DNS Settings get lost after install" bug.
Everything seems to work fine so far.

[Gostev]

And literally in parallel a have someone complaining to me how incredibly hard it is to find the "Veeam Infrastructure Appliance" ISO download on our web site and how well it is hidden in the deepest and the darkest place of "Additional downloads" section go figure!

[nick_zadara]

I also ran into this issue of the 443 web UI not loading and seeing the NGINX HTTP Server Test Page, yet the UI for 10443 worked ok, after performing an update which required a reboot of the VM.
And then the same issue from a Windows VM attempting to connect to the Software Appliance from the VBR Console - "Failed to connect to the backup server: Failed to connect to identity service (https://v13-appliance/) - NotFound"
For me it turned out to be because I have 2 vNICs assigned to the VBR Software Appliance VM.
To resolve it I needed to enable the 2nd vNIC as it didn't automatically start on bootup.
I then remembered, I had to manually enable it during the initial ISO install of the VM.
I believe the symptom is related to DNS, but the root cause was the disabled vNIC.
So I think there may be a bug with an additional vNIC not being enabled by default at bootup.