NCP keeps restarting with the following error: Resource could not be found on backend <...> for GET policy/api/v1/infra/ip-blocks/<ip_block_name>tile.

Products

• VMware NSX
• VMware NSX Networking

Issue/Introduction

NCP is not operational, cannot deploy or scale any application in TAS.

NCP keeps restarting with the following error: Resource could not be found on backend <...> for GET policy/api/v1/infra/ip-blocks/<ip_block_name>tile.

Symptoms: Customer configured IP_BlocK_NAME "IP_BlocK_NAME" on TPCF for NCP to communicate with NSX-T manager. However the request `GET /infra/ip-blocks/IP_BlocK_NAME` keeps failing not matter that the ip-block does exist on NSX-T. Once replace the ip-block from name to UUID, it works. According to NSX-T API document, both name and UUID are supported.

NSX 4479 SYSTEM [nsx@6876 audit="true" comp="nsx-manager" level="INFO" subcomp="manager"] UserName:'NSX_FQDN@local' ModuleName:'common-services' Operation:'GET@/api/v1/infra/ip-blocks/IP_BlocK_NAME' Operation status: 'failure' Error: The requested object : /infra/ip-blocks/IP_BlocK_NAME could not be found. Object identifiers are case sensitive.

 

 

Environment

 Found on NSX-T with NCP 4.1.x

Cause

Starting with NCP 4.1.x, the "domain group" created for the TAS foundation has been changed from dynamic to static and is built using container IP blocks' CIDRs. However, it expects container IP blocks ids, and therefore the creation of the domain group will fail if the container IP block was specified by name. As this is a startup operation, NCP will fail to initialise in this case.

Resolution

Issue has already been fixed in NCP 4.1.2.

 

Workaround:

Update the NSX tile to specify the IP block id rather than its name and deploy to update NCP configuration.