Symptoms

VxRail Manager runs health-checks using VxVerify, one of which is vc_api. This check looks for errors on API connections to vCenter (VC), using the saved management credentials in VxRail Manager.If the test passes, it returns the following event, indicating the status of the VC API connection attempts: Test Result Result Code Result InterpretationPass0API connections to vCenter were successful.Warning1Not applicable for this testFailure2Not applicable for this testCritical3VxRM to VC API connection failed Each test that passes is not listed in the summary report, for ease of reading.An example of the health-check output is shown below: #========================#======#=========#====================================================================#==============# | Hostname / Category |Status Dell_KB | Warnings or Failures, unless tests Passed ; Product S.N. | #========================#======#=========#====================================================================#==============# | VxRM | Critical 201234 | vc_api: VC API connection failed. See vxv.log. |

Cause

A critical failure from this health-check means that the API connection could not be established. A critical response is more severe than a standard test failure, because in this case it means that other health-checks (which use API), cannot be run.

Resolution

Critical errors should be fixed before warning or failure results and then the tests should be rerun. For example, the results of any tests that use the VC API would be unreliable if the API connection has a fault.Examples of critical responses and the typical actions to take are listed below: Saved management password rejected on vCenter If this health-check responds with the above event, the management username and password saved in VxRail Manager (in 7.x this would be in the lockbox), are rejected in vCenter. The following event would appear in vxv.log, showing the failing vCenter management user (such as vxrm@localos), attempting to connect to vCenter.For example from vxv.log: [vc_connection] Management user and password from the VxRM lockbox was rejected by vCenter If there is a problem with the saved credentials, such as these no longer matching the vCenter management password, then the saved credentials must be updated to match this vCenter user's password. See Solve for the procedures to do this: Dell Technologies SolVe Online: VxRail Procedures > Miscellaneous > 'How To' Procedures > Change VxRail Passwords > Select Version > VxRail External Management User Password VC API connection failed. See vxv.log. The above result is returned for failing API connections that do not have an Invalid Login response from vCenter. These could be the result of an incorrect vCenter name being stored in VxRail Manager (in 7.x this would come from the config_service entry for vcenter_host).This issue also occurs when the vCenter FQDN cannot be resolved from VxRail Manager. An additional error can be logged in the VxVerify report for dnslookup. See article Health-check fails for dnslookup.To get details of this event, the vxv.log should be reviewed. See article Troubleshooting when VxVerify reports an error (Log in as a registered Dell Support user may be required to view this article.) Unsupported characters in VC management password The use of some special characters, permitted by vCenter, cause issues for API. If bad characters in the management password are the cause of the API connection failure, then there should also be a health-check failure for test 'vc_pw_char'. VC MOB API connection failed VxRail Manager could not reach the vCenter MOB API. Check that the following link is reachable, which gives a login prompt (use the vCenter management login for VxRM, not administrator). https://<VC_FQDN>/mob Logging in successfully would show: Managed Object Type: ManagedObjectReference:ServiceInstance If the MOB connection fails, it could be because the MOB is disabled, which could have been done for security reasons. Check in VCenter /etc/vmware-rhttpproxy/endpoints.conf.d/vpxd-rhttpproxy-endpoint.confIf the /mob line is commented out (with #), then this disables the MOB. To enable MOB: Uncomment the line beginning with /mob in file: /etc/vmware-rhttpproxy/endpoints.conf.d/vpxd-rhttpproxy-endpoint.conf Restart vmware-rhttpproxy service-control --restart vmware-rhttpproxy The MOB can be disabled again after the upgrade by reverting the change above.

Support Cases