NFS Datastore Intermittently Flaps and VM Migration Fails After ESXi and ONTAP Update

After upgrading both ESXi hosts to 8.0.3 (Build 24674464) and the ONTAP Select cluster to 9.16.1P8, the following issues were observed:

• The NFS datastore (e.g., vol1) repeatedly becomes inaccessible (“flapping”) on one ESXi host (esx004), though other NFS datastores remain healthy.
• The compute vmotion of the VMs hosted on the datastore from the ONTAP select is still failing.
• The vmotion stuck at 69% for a long period and failed eventually.
• Engaged VMware/Broadcom support for joint troubleshooting; no ONTAP-side errors were detected.
• VM kernel log reports all paths down, but there were no identifiable error from the logs.

2025-11-07T15:47:19.025Z vmkernel: StorageApdHandler: APD start for 0x4320a9e09c20 [badf26a0-5da34d75]
2025-11-07T15:47:20.033Z hostd: Datastore 10.130.29.41:/vol1 has accessibility of false due to AllPathsDown_Start
2025-11-07T15:48:43.028Z vmkwarning: NFS: Lost connection to the server 10.130.29.41 mount point /vol1, mounted as badf26a0-5da34d75-0000-000000000000 ("vol1")
2025-11-07T15:49:01.957Z vmkernel: NFS: Restored connection to the server 10.130.29.41 mount point /vol1, mounted as badf26a0-5da34d75-0000-000000000000 ("vol1")
025-11-07T15:49:01.958Z In(14) vobd[2097764]:  [APDCorrelator] 347464964047us: [esx.clear.storage.apd.exit] Device or filesystem with identifier [badf26a0-5da34d75] has exited the All Paths Down state.
2025-11-07T15:49:55.028Z In(14) vobd[2097764]:  [APDCorrelator] 347513978844us: [vob.storage.apd.start] Device or filesystem with identifier [badf26a0-5da34d75] has entered the All Paths Down state.
2025-11-07T15:49:55.028Z In(14) vobd[2097764]:  [APDCorrelator] 347518034540us: [esx.problem.storage.apd.start] Device or filesystem with identifier [badf26a0-5da34d75] has entered the All Paths Down state.