silk-datastore-syncer job fails after upgrade to TAS 4.x when using NCP

Products

• VMware Tanzu Application Service for VMs

Issue/Introduction

After upgrade to any TAS 4.x version below 4.0.7 LTS-T in a vSphere environment with NSX-T NCP as CNI, Diego cells show in failing state due to Silk-Datastore-Syncer job failing with error:

{"timestamp":"202x-xx-xxT16:04:05.258832013Z","level":"info","source":"cfnetworking.silk-datastore-syncer","message":"cfnetworking.silk-datastore-syncer.properties","data":{"gardenAddr":"/var/vcap/data/garden/garden.sock","gardenNetwork":"unix","interval":30,"silkFile":"/var/vcap/data/container-metadata/store.json","silkFileGroup":"vcap","silkFileOwner":"vcap"}}

goroutine 1 [running]:
main.main()
        /var/vcap/data/compile/silk-datastore-syncer/src/code.cloudfoundry.org/silk-datastore-syncer/main.go:106 +0xf52
panic: assignment to entry in nil map

Environment

Product Version: 4.0

Resolution

To resolve this problem, upgrade to TAS v4.0.7 LTS or higher.
The fix is included in silk release v3.34.0