CUCM v15 Installation NTP inaccessible

[Pavan Manjunath]

Hi,

I'm trying to install CUCM v15.0.1.11901 from VMware ESXi v7.0.3. When i enter the NTP address it throws this error, "the ntp server(s) are mistyped, inaccessible or unreliable. Verify that valid NTP server names or ip addresses were entered, that are running NTPv4 at stratum 6 or less, and the port 123 is not blocked by the firewall" 

The same NTP address, I have given for CUCM v14.0.1.12900. The NTP address worked and installation setup successfully completed. 

I'm not sure why it's throwing inaccessible for v15. Need help to fix this.

I did checked with my local IT team and confirmed that NTP is accessible and no firewall block is found.

CUCM 15 Configuration Details

Guest OS: Red Hat Enterprise Linux 7 (64-bit)
Compatibility: ESXi 7.0 U2 virtual machine
VMWareTools: Yes
CPUs: 2
Memory: 10 GB
Hardisk 1: 110GB

[Pavan Manjunath]

@Roger Kallberg I would appreciate your assistance with this issue, as I reviewed the document but couldn't find any helpful information.

[Roger Kallberg]

Guest OS should not be RHEL. It should be Other Linux 64-bit. Have you created the VM with the latest OVA for version 15?

[Brad Magnani]

You may want to double check with your NTP team if the NTP server is running on Windows.

https://www.cisco.com/c/en/us/support/docs/unified-communications/unified-communications-manager-callmanager/215537-network-time-protocol-ntp-on-cucm.html

The Windows NTP server is not supported for CUCM; however, other types such as Linux NTP sources, Cisco IOS® NTP sources, and Nexus OS NTP sources are acceptable.

Although other Cisco solutions can utilize Windows Servers for the NTP solution, UC Solutions such as Call Manager, Cisco Unity, and Instant Messaging and Presence, are unable to do so and require either a Linux-based or Cisco IOS®-based NTP solution.

This is because Windows Time Services often uses SNTP which Linux systems have difficulty synchronizing with.

SRND also mentions this:
https://www.cisco.com/c/en/us/td/docs/voice_ip_comm/cucm/srnd/collab12/collab12/netstruc.html?

Using Windows Time Services as an NTP server is not recommended or supported because Windows Time Services often use Simple Network Time Protocol (SNTP), and Cisco Unified CM cannot successfully synchronize with SNTP.

The external NTP server specified for the primary node should be NTP v4 (version 4) to avoid potential compatibility, accuracy, and network jitter problems. External NTP servers must be NTP v4 if IPv6 addressing is used

If it's running on a supported platform, it could be an issue with network connectivity from this VM's vNIC/gateway, etc.  You'd need to run a packet capture upstream of the ESXi host to confirm NTP traffic is leaving the VM towards the NTP server and NTP is responding back. 

[Pavan Manjunath]

@Roger Kallberg Thanks for your reply

The ova verison which we are using is Bootable_UCSInstall_UCOS_15.0.1.11901-2.sha512.iso --> demo version
Could you confirm if it is the latest version or let me know the latest version

[Brad Magnani]

That's the ISO, not to be confused for the OVA.  The OVA sets all the necessary specs prior to installing the VM, which is located here: https://software.cisco.com/download/home/286331940/type/283088407/release/15

ReadMe: https://www.cisco.com/web/software/283088407/166989/cucm_15.0.ova.README.txt

Assuming you deployed the correct OVA, the ISO software version you mention is fine - the NTP issue is not due to the software version, see my comments above for how to further isolate the issue.

[Roger Kallberg]

That not an OVA, it’s an ISO. To create the VM you’ll use an OVA, that you download from Cisco and then you’ll use the ISO to have the VM boot from to install the product, for example Communication Manager.

[Roger Kallberg]

That is btw not the latest version of the ISO, you’re using SU1 version. There is a SU2 version released.

[Nithin Eluvathingal]

You mentioned the same IP was used for version 14. Have you confirmed it was synced or showing as synchronized in 14?

I installed 15FCS today, and during installation, the same IP I used on PCD, which was synced, appeared inaccessible. When I selected the option to proceed, the installation completed, and post-installation, the NTP was synced.

In your case, ensure the IP used with 14 is synchronized. If not, the IP may be running NTP service, which could be the issue.

A few points to clarify:

• OVA is a template.
• Bootable ISO is the media mounted on the OVA for installation.
• There is no DEMO OVA or DEMO ISO.
• UC applications provide a 90-day evaluation by default; no separate DEMO version exists.

2.12.0.0

[Pavan Manjunath]

@Roger Kallberg @Brad Magnani @Nithin Eluvathingal 

Thank you for your help and support. I followed all your suggestions and attempted to reinstall using the OVA file cucm_15.0_vmv17_v1.2.sha512.ova, but it appears to be corrupted.

I manually configured the following details:

• Guest OS: Other Linux 64-bit

• Compatibility: ESXi 7.0 U2 virtual machine

• VMware Tools: Enabled

• CPUs: 2

• Memory: 12 GB

• Hard Disk 1: 110 GB

I also selected the bootable ISO from datastore1. However, the installation still failed, displaying the "NTP inaccessible" error.

For additional context, I ran two commands on CUCM v14, where the NTP was functional, but the ntp status command showed it as unsynchronized:

• Command: utils ntp status Output: Unsynchronized (polling server every 8 seconds)

• Command: utils diagnose test Output: Diagnostic tests passed for disk space, service manager, Tomcat, validate network, RAID, and NTP-related checks (reachability, clock drift, stratum).

I’m unsure how to synchronize the NTP server on CUCM v14 to replicate the same configuration for CUCM v15. Notably, this is my second attempt to install CUCM v15. Three months ago, the same NTP IP worked without issues in v14 & v15. Now it fails despite successful ping tests, and local IT has asked me to review the v15 configuration since NTP works in v14.

Your insights and guidance would be greatly appreciated. Thank you

[Roger Kallberg]

You should not manually configure the VM. You should create the VM by using the deploy with OVF option and using the OVA.

[Pavan Manjunath]

@Roger Kallberg Could you provide the latest OVA file for version 15, as the one I have is corrupted?

[Roger Kallberg]

How do you know it’s corrupt? How are you using the OVA, can you please outline in detail the steps you took?

[Pavan Manjunath]

@Roger Kallberg 

I obtained the cucm_15.0_vmv17_v1.2.sha512.ova file from my teammate and downloaded it to my local system.

1. Logged into VMware ESXi.

2. Clicked on Create/Register VM.

3. Selected the creation type: Deploy a virtual machine from an OVF or OVA file, then clicked Next.

4. In the Select OVF and VMDK files step:

   • Provided the virtual machine name.

   • Selected the OVA file from my local system.

   • Clicked Next.

5. Chose storage: datastore1, then clicked Next. The OVA extraction began, which might take some time.

6. Configured the deployment options:

   • Network mappings: Selected eth0, which is the network provided by the IT team (not the VM network).

   • Deployment type: Medium UCM.

   • Disk provision: Thin.

   • Power ON automatically: Enabled (checked the checkbox).

   • Clicked Next.

In the Ready to complete step, where the settings are reviewed before finalizing the wizard, the following error appeared: "There was an error creating the import specification from the OVF file. Review your settings selection before finishing the wizard."

Let me know if further clarification is needed!

 

[Roger Kallberg]

Thank for the specific details. What you’re describing doesn’t look the same as I recall from when doing this a few months ago. Tomorrow when I’m at work I’ll check on this in our own system and provide you information on how it looks there.