Site pairing fails with error 'Site-Pair already exists between the sites' in HCX
search cancel

Site pairing fails with error 'Site-Pair already exists between the sites' in HCX

book

Article ID: 419048

calendar_today

Updated On:

Products

VMware HCX

Issue/Introduction

  • From HCX Standalone UI -> Site Pairs.  Attempting to add a site pairing does not result in a site pairing being displayed or having been created.
    Note: The same error may be experienced: "Site-Pair already exists between the sites", if attempting to use the  API to reconnect, for more information: Reconnecting an HCX Site Pairing via API fails with the error: "Site-Pair already exists between the sites"
  • Site Pair creation is being attempted, and it's failing with the error: "Site-Pair already exists between the sites"
  • The following error is observed in /common/logs/admin/web.log
    <timestamps> UTC [https-jsse-nio-8443-exec-4, Ent: HybridityAdmin, , TxId: TxId: ########-####-####-####-########] ERROR c.v.v.h.api.registration.CloudConfig- Unable to addCloud: Site-Pair already exists between the sites
    com.vmware.vchs.hybridity.api.exception.BadRequestException: Site-Pair already exists between the sites

Environment

VMware HCX

Cause

This issue occurs when residual entries remain in the HCX Manager database, creating a mismatch between the database state and the UI. Because the site pair does not  exist in the UI but the database still contains stale records, the API cannot communicate with the remote site to perform a standard cleanup. This forces the need for manual database cleanup to remove the orphaned entries before a new site pair can be established.

Common scenarios that lead to this condition include:

  • Site pair deleted from only one side of the connection
  • HCX Manager restoration from backup after site pair was removed on the peer site
  • Network connectivity loss during site pair removal/creation
  • Incomplete or interrupted site pair deletion/creation operations

Resolution

Resolution :Fixed in release 4.11.4 and higher. See Download Broadcom products and software for steps to download this release.
Now, if there are site-pair records already on the source side, but not listed in UI, then new CREATE site-pair between the same pair of sites will clear the old records on source and create new site-pair records. If there are site-pair records already on target side and if a new incoming site-pair creation request comes to the target then the existing records are deleted and new site pair records are created on the target side.

Workaround:

Broadcom Support provides two primary methods to work around this condition.

Option 1: Manual Database Cleanup (Engineering Engagement)

This method is preferred if redeploying the appliance is not feasible due to existing configurations or service uptime requirements.

  1. Collect HCX Technical Support bundles from both the source (Connector) and target (Cloud) managers. Ensure the "Include Database" option is selected.
  2. Open a support case with Broadcom Support.
  3. Support will engage Engineering to provide a site-specific SQL script to identify and remove the orphaned site pair entries from the database..
    • Note: Do not attempt to modify the database manually without an Engineering-provided script.

Option 2: Appliance Redeployment (Self-Service)

This is often the fastest resolution path if the issue is localized to the HCX Connector and a maintenance window is available.

  1. Information Gathering: (Pre-flight): Take screenshots or document your existing HCX Connector configurations, including IP addresses, DNS, NTP, and Compute Profile settings (Network Profiles and Uplinks).
  2. Power off the corrupted HCX Connector appliance, rename it, and do a storage vmotion to apply the new name, then remove from inventory.
  3. Deploy a new HCX Connector OVA using the same IP and hostname.
  4. Do NOT restore the configuration backup file. Manually reconfigure the baseline system settings, vCenter/NSX manager registrations, and Compute Profile.
  5. Test Connectivity: Before executing the pairing wizard, verify bidirectional communication via ping [Target_IP_FQDN] and curl -k -v https://[Target_IP_FQDN]:443 ssh’d into the newly deployed manager.
  6. Pairing: Re-attempt the site pairing using the fresh database.

Why the IP-Based Pairing Step Matters

When using an IP address for site pairing due to a lack of cloud-side DNS,(connectivity fails on FQDN but passes on IP) the curl -k (insecure) flag in Step 5 is critical. It allows testing the raw TCP port 443 handshake and verify that the target web server responds, ignoring any SSL certificate name mismatch errors that naturally occur when navigating directly to an IP address rather than a validated FQDN.

Additional Information

Reconnecting an HCX Site Pairing via API fails with the error: "Site-Pair already exists between the sites"

If contacting Broadcom support about this issue,  provide the following:

  • HCX Manager Connector logs (including the database).
  • HCX Manager Cloud logs (including the database).
  • HCX User Guide :Site Pairing.
Powered by Wolken Software