My Oracle Support Banner

OSMS agent service fails to start on instances with more than one VNIC (Doc ID 2750767.1)

Last updated on MARCH 24, 2021

Applies to:

Oracle Cloud Infrastructure - Version N/A and later
Information in this document applies to any platform.

Symptoms

OSMS agent service fails to start on instances with more than one VNIC. Attempts to start the agent service manually on Windows fails as well. You may see associated entries in the Windows event log stating the service timed-out or took too long to respond.

For illustration purposes, the routing table from an instance with two VNICs is provided as follows. The primary interface on the instance has the IP address 10.x.0.5 and the secondary interface has the IP address 10.x.0.21

1. Routing table as-is  (from output of command "route print") with traffic to 169.254.169.254 going through the secondary interface with IP address 10.x.0.21:

Network Destination          Netmask    Gateway    Interface  Metric
     169.254.0.0         255.255.0.0    On-link    10.x.0.5       16       
 169.254.169.254     255.255.255.255   10.x.0.1   10.x.0.21       16
 169.254.255.255     255.255.255.255    On-link    10.x.0.5      271

2. Routing table with the secondary interface temporarily disabled (from within Windows) which resolves the issue (restarting OracleCloudAgent service causes the OSMS agent to start):

Network Destination          Netmask    Gateway    Interface  Metric
169.254.0.0              255.255.0.0    On-link    10.x.0.5       16
169.254.255.255      255.255.255.255    On-link    10.x.0.5      271


3. You may see lines similar to the following in the OSMS agent.log file (e.g., C:\Windows\ServiceProfiles\OCAOSMS\AppData\Local\osms\agent.log):
2021-02-05 11:54:51,879 - ERROR - Fail to create instance principal signer. Error message: 'NoneType' object has no attribute 'get'
2021-02-05 11:57:43,913 - ERROR - Fail to create instance principal signer. Error message: 'NoneType' object has no attribute 'get'
2021-02-05 11:57:43,913 - WARNING - No regionInfo key detected. Defaulting to oraclecloud domain.
2021-02-05 11:57:43,913 - INFO - Endpoint: https://ingestion.osms.None.oci.oraclecloud.com

4. You may see lines similar to the following in the OCA agent.log file (e.g., C:\Windows\ServiceProfiles\OCA\AppData\Local\OracleCloudAgent\agent.log):
2021/02/05 12:36:32.012305 health.go:103: health check of plugin:[unifiedmonitoring], desiredState:[Unknown], currentState:[notStarted], status:[yet to start] version:[] err:[<nil>]
2021/02/05 12:36:32.013292 instancemetadata_client.go:61: failed to get metadata from http://169.254.169.254/opc/v2/instance/ , status 404 Not Found
2021/02/05 12:36:32.014302 instancemetadata_client.go:75: failed to get metadata from http://169.254.169.254/opc/v1/instance/ , status 404 Not Found
2021/02/05 12:36:32.014302 status.go:116: plugin status update err: failed to get metadata from http://169.254.169.254/opc/v1/instance/ , status 404 Not Found

Changes

 Addition of a secondary VNIC to the instance

Cause

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


In this Document
Symptoms
Changes
Cause
Solution
 Applies to:


My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.