update real allocated size of a volume, after it is created

[silentred]

When following the quick start document, the LVS type volume cannot be succussfully created.

To make sture that a storagepool exists

kubectl -n obnvmf get storagepool
NAME                                               IP              HOSTNAME                                           STORAGE   FREE    STATUS   AGE
ip-172-26-14-129.ap-northeast-1.compute.internal   172.26.14.129   ip-172-26-14-129.ap-northeast-1.compute.internal   998Mi     998Mi   ready    54m

Show the detailed info of the storagepool

kubectl -n obnvmf get storagepool ip-172-26-14-129.ap-northeast-1.compute.internal -oyaml
apiVersion: volume.antstor.alipay.com/v1
kind: StoragePool
metadata:
  annotations:
    obnvmf/hostnqn: nqn.2021-03.com.alipay.host:uuid:ip-172-26-14-129.ap-northeast-1.compute.internal
    obnvmf/tgt-version: ""
  creationTimestamp: "2023-11-29T08:06:30Z"
  finalizers:
  - antstor.alipay.com/in-state
  generation: 1
  labels:
    obnvmf/local-storage-bytes: "1046478848"
    obnvmf/node-sn: ip-172-26-14-129.ap-northeast-1.compute.internal
  name: ip-172-26-14-129.ap-northeast-1.compute.internal
  namespace: obnvmf
  resourceVersion: "5713"
  uid: 20d1c020-c53a-45b2-8753-aa2b1b06bd75
spec:
  addresses:
  - address: 172.26.14.129
    type: InternalIP
  kernelLvm: {}
  nodeInfo:
    hostname: ip-172-26-14-129.ap-northeast-1.compute.internal
    id: ip-172-26-14-129.ap-northeast-1.compute.internal
    ip: 172.26.14.129
    labels:
      beta.kubernetes.io/arch: amd64
      beta.kubernetes.io/os: linux
      custom.k8s.alipay.com/obnvmf-host: "true"
      kubernetes.io/arch: amd64
      kubernetes.io/hostname: ip-172-26-14-129.ap-northeast-1.compute.internal
      kubernetes.io/os: linux
      node-role.kubernetes.io/control-plane: ""
      node.kubernetes.io/exclude-from-external-load-balancers: ""
  spdkLVStore:
    baseBdev: test-aio-bdev
    blockSize: 512
    bytes: 1046478848
    clusterSize: 2097152
    name: aio-lvs
    totalDataClusters: 499
    uuid: fa6ac6bd-89cd-4b03-a365-44ae7ff3c7e8
status:
  capacity:
    storage: 998Mi
  conditions:
  - status: OK
    type: Spdk
  status: ready
  vgFreeSize: 998Mi

the disk-agent daemonset pod has logs

E1129 08:13:18.190071       1 sync_loop.go:128] obnvmf/pvc-9c1af30c-5156-4597-9ae8-67e68f2e2137Code=-19 Msg=No such device
I1129 08:13:32.689461       1 heartbeat.go:37] Heartbeat is ok
I1129 08:13:48.192599       1 sync_loop.go:102] process key obnvmf/pvc-9c1af30c-5156-4597-9ae8-67e68f2e2137
I1129 08:13:48.196292       1 sync_volume.go:96] syncing volume pvc-9c1af30c-5156-4597-9ae8-67e68f2e2137 to create logic vol and spdk target
I1129 08:13:48.196313       1 sync_volume.go:167] volume pvc-9c1af30c-5156-4597-9ae8-67e68f2e2137 is not in expansion
2023/11/29 08:13:48 client.go:54: {"jsonrpc":"2.0","method":"bdev_get_bdevs","id":26,"params":{"name":"aio-lvs/pvc-9c1af30c-5156-4597-9ae8-67e68f2e2137"}}
2023/11/29 08:13:48 client.go:65:
E1129 08:13:48.202368       1 service.go:119] Code=-19 Msg=No such device
E1129 08:13:48.202384       1 lvs.go:94] Code=-19 Msg=No such device
E1129 08:13:48.202392       1 sync_volume.go:106] Code=-19 Msg=No such device
E1129 08:13:48.202402       1 sync_loop.go:120] key obnvmf/pvc-9c1af30c-5156-4597-9ae8-67e68f2e2137 has retry too many times 6, delay it for 30s

The problem is at this line:

liteio/pkg/agent/sync/sync_volume.go

Line 649 in ebb490b

vol, err = vs.poolService.PoolEngine().GetVolume(lvName)

If the PoolEngine is SpdkLVS and the Lvol is not created, GetVolume returns an error. This causes the volume hangs in creating status.