Azure HA vSocket Troubleshooting

Overview

This article provides insights into common Azure vSocket HA deployment issues and offers troubleshooting steps to resolve them. This guide aims to assist in identifying and addressing potential hurdles during and after deploying the HA solution either via HA script or Marketplace.

Symptoms

When deploying Azure vSocket HA, you might encounter the following symptoms:

  • HA Script Failure
    • Failure in executing the create_ha_settings script in manual deployment.
    • Issues with secondary vSocket deployment via Marketplace.
  • HA Failover Failure
    • Failed HA API tests from the Socket WebUI.
    • Failed HA failover resulting in traffic not being forwarded to the secondary vSocket.
  • HA Status Not Ready
    • CMA shows that the HA status of the Site is Not Ready

Possible Causes

The most frequent causes of HA deployment failures include:

  • Use of non-public DNS in Azure.
  • Management interface lacking internet access.
  • Insufficient Azure account permissions.
  • Restrictive Security Group and Routing settings in Azure.
  • Failure to assign the floating IP address to the LAN interface.
  • LAN connectivity issues.

Troubleshooting the Issue

Important

IMPORTANT: Before starting to troubleshoot, ensure to verify all prerequisites for Azure HA vSocket deployment. See Configuring High Availability (HA) for Azure vSockets and Deploying Azure vSockets from the Marketplace

Troubleshooting HA Script Failure

The Azure HA script (create_ha_settings) and secondary vSocket deployment via Marketplace validate that the Azure subscription has two valid vSockets and then assign the Roles that create the HA and failover mechanism. If the script run fails, follow these troubleshooting steps:

Checking Activity logs

  • In Azure, Activity Logs store all the events that happened inside each Azure resource. Review these logs if the deployment isn't successful and one of the roles is not assigned. Browse to the VM or the NIC and select the activity log

Checking Azure Naming Restrictions 

  • When entering the vSocket Name in the script, make sure that the name doesn't include spaces or restricted characters as explained in Naming rules and restrictions.
  • If a naming issue occurs during deployment, the following error will be shown in the Error log
    The value of parameter disk.name is invalid. (Code: InvalidParameter, Target: disk.name)

Checking Azure DNS Configuration

  • Make sure that the default Azure DNS is configured for both the VNET and the associated NICs. If the default DNS isn't configured in Azure in both the VNET and NIC, Role creation will fail.
  • To check what the Azure DNS configuration is, see Fixing DNS configuration issues 

Checking Azure Permissions

  • To successfully run the HA script, make sure that the Azure user has owner permissions. Go to Resource Group > Access Control IAM > Vier My Access, and verify that the user account is set to Owner or higher role.

Verifying Azure Role Assignment 

  • The HA script creates the Role that is stored in Azure Managed Identities. 
  • The script assigns this Role to the different virtual resources that are attached to the vSocket. The components in the Azure infrastructure that use the Role are:

    • LAN Network Interface (NIC) for each vSocket
    • The LAN subnet associated with the LAN NICs
    • Both vSocket VMs
  • Only one user-assigned role should be assigned to each resource. If there is a policy adding system-assigned identities in Azure, the vSockets must be excluded from it.
  • The role assignment for the NICs can be checked under Access control > Role Assignments and should be assigned for both Primary and Secondary LAN NICs.
  • The role assignment for the LAN subnet can be checked under VNET > Subnet, then select the LAN subnet and click Manage users > Role Assignments
  • For each vSocket VM, the identity role can be checked under Identity > User assigned as you can see in the screenshot below. 

Re-running the HA script

  • As a last resort, the HA script (create_ha_settings) can be re-run once the previous steps have been checked.
  • Make sure to renew the Azure token and remove the Azure Managed Identity if it was created during the first script run.

 

Troubleshooting HA Failover Failure

If the HA script runs successfully but the vSocket HA failover doesn't occur as expected (e.g. traffic isn't routed to the secondary vSocket), follow these steps:

Running HA API Test

  • From the vSocket webUI, run the API Test tool from both vSockets which validates that the API Call to Azure can be made successfully. Any errors with permissions or Floating IP assignments can be seen here.

Checking the Activity log

  • In Azure, Activity Logs store all the events that happened inside each Azure resource. Review these logs to identify if the floating IP failed to be pushed to the LAN NIC or if the API Test isn't successful. Browse to the NIC and select the activity log

Pinging the Floating IP

  • From the vSocket WebUI, utilize the Ping tool, select the LAN interface, and ping the Floating IP address. If this test isn't successful, continue with Verifying the Floating IP assignment

Verifying the Floating IP assignment

  • To route traffic to the master vSocket, Azure assigns the Floating IP to the LAN NIC of the current master vSocket. Go to the primary vSocket VM LAN NIC > IP Configuration and verify that the Floating IP exists as "Secondary". If not, continue with the next steps.

Verifying Azure Role Assignment 

  • Run the steps provided in Verifying Azure Role Assignment to confirm that the identity role listed in the resource group is assigned to the LAN NICs, LAN subnet, and both vSocket VMs.

Checking that the Management interface has DNS and internet access

  • Verify that the management interface has internet access and can connect to the configured DNS server.
  • Check DNS resolution for management.azure.com from the Azure portal. The HA API Call uses this FQDN.

    • Go to the Virtual machine > vSocket > Run command > RunShellScript
    • Enter dig management.azure.com in the text box
    • Click Run
  • The dig output will be displayed in the portal with a DNS response.

  • If there's no DNS resolution, see Fixing DNS configuration issues.
  • From the same page, try to reach any internet resource to confirm internet access. For example, ping -c 4 8.8.8.8 If the ping is not successful, continue with the next steps.

Check if the Network Security Group is blocking outbound traffic.

  • A quick way to check is to go to the management network interface in Azure and click "Effective security rules" at the bottom left side of the screen.
  • The screenshot below shows no NSG assigned, so outbound traffic is not blocked.

Check the MGMT Interface Routing for Internet Traffic.

  • In case MGMT interface traffic is routed through a third-party firewall in Azure, check that UDP/53 and TCP/443 outbound connections are allowed.
  • The route table can be reviewed on the management interface page in Azure by clicking the "Effective routes" option.
  • The screenshot below shows the route for internet traffic using the Internet as the "Next Hop Type", so a firewall is not blocking the traffic.

Checking the Routing Table's Next Hop

  • Confirm that the LAN Routing table is pointing at the Floating IP. Change the Next-hop IP address accordingly if needed.

Troubleshooting HA Status Not Ready

If CMA shows that the HA Status is Not Ready and both vSockets are up and running, both vSockets will take the Master role (split-brain scenario). There could be two associated problems:

  • Both vSockets are running different firmware versions
  • HA Keepalive messages do not reach the secondary vSocket

It is recommended to check both vSockets WebUI pages to confirm the HA status of each one of them. A split-brain scenario will manifest if both primary and secondary vSockets are in a Master role. The webUI will show the current role on the top of the main Monitoring page.

Checking Firmware versions

To satisfy the compatible version criteria, both vSockets must be running the same MAJOR version, e.g. v17.xx.yy or v18.xx.yy. vSockets perform an initial upgrade after first being deployed. If one of the vSockets fails to upgrade, this issue must be troubleshot. Submit a Support ticket to report this issue.

Checking HA Keepalives

Keepalive packets use port UDP/20480 for Azure vSocket and will be sent only from the Master vSocket to the Standby vSocket. The split-brain condition occurs when both vSockets have the Master role, which can happen due to LAN connectivity issues between the vSockets that create a situation in which the HA Keepalive messages do not reach the secondary vSocket. 

Run the following checks to confirm LAN connectivity:

  • Check if the Network Security Group is blocking port UDP/20480. A quick way to check NSG rules is to go to each LAN network interface in Azure and click "Effective security rules" at the bottom left side of the screen.
  • Confirm that both LAN interfaces are associated with the same LAN subnet.
  • Run a packet capture from the WebUI of both vSockets and identify if the keepalives sent by the primary are being received by the secondary vSocket.

Resolving Discovered Issues

Renewing Azure token

  • If Azure Cloud Shell is used to deploy the HA script, open a new session and re-authenticate. This will renew the token used for querying the API.

Fixing DNS configuration issues

  • To fix the Azure DNS configuration and set it to the default value, go to Virtual network > DNS servers and to Network Interface > DNS Servers, and ensure that you're using the Default option. Shut down the VM to make any DNS-related changes and then turn it back up.

Unregistering and Redeploying an Azure vSocket

  • If after following all the above troubleshooting steps, the HA script or HA failover continues to fail, it is possible to unregister and redeploy one or both vSockets. See Redeploy High Availability vSocket Sites
  • It is important to follow the guidelines and remove the Virtual Machine, network interfaces, associated public IPs, and Managed Identity before redeploying a vSocket.
  • If only the primary vSocket instance is re-deployed, you must run the dedicated HA script (create_ha_settings) to bind both vSocket instances for HA.

 

Raising cases to Cato Support

Submit a Support ticket with the results of the above troubleshooting steps. Please include the following information in the ticket:

  • A clear description of the issue including any error messages.
  • DNS test results for management.azure.com
  • API Test results.
  • Screenshots of the assigned floating IP and configured identity roles.
  • Azure activity log screenshots including any errors found.

 

Was this article helpful?

1 out of 2 found this helpful

3 comments

Add your comment