PowerScale: nodes panic with stack: Assertion Failure kernel:pdm_update_membership+0x10b2

Multiple nodes panic with stack:
 

2022-03-03T14:40:11.340255-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: panic @ time 1646335137.652, thread 0xfffffe8ffa3b8080: Assertion Failure
2022-03-03T14:40:11.340297-05:00 <5.3> (id12) syslogd: sendto: Network is unreachable
2022-03-03T14:40:11.340317-05:00 <5.3>  syslogd: last message repeated 1 times
2022-03-03T14:40:11.340322-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: time = 1646335137
2022-03-03T14:40:11.340334-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: cpuid = 15, TSC = 0xc868113bb278f
2022-03-03T14:40:11.340340-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: Panic occurred in module kernel loaded at 0xffffffff80200000:
2022-03-03T14:40:11.340347-05:00 <0.7> (id12) /boot/kernel.amd64/kernel:
2022-03-03T14:40:11.340353-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: Stack: --------------------------------------------------
2022-03-03T14:40:11.340359-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:isi_assert_halt+0x2e
2022-03-03T14:40:11.340365-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:pdm_update_membership+0x10b2
2022-03-03T14:40:11.340371-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:pdm_get_painted_domids+0x568
2022-03-03T14:40:11.340377-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:pdm_member_generate_entry+0xda
2022-03-03T14:40:11.340383-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:pdm_member_get_membership_entry+0x1b7
2022-03-03T14:40:11.340390-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:pdm_member_init_operation+0x16d
2022-03-03T14:40:11.340396-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:ifm_init_operation+0xb7
2022-03-03T14:40:11.340402-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:txn_i_include_vnode_to_list+0xc0
--
2022-03-03T14:40:11.340440-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:VOP_RENAME_APV+0x9b
2022-03-03T14:40:11.340446-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:isi_kern_renameat+0x37e
2022-03-03T14:40:11.340452-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:sys_enc_renameat+0x27
2022-03-03T14:40:11.340458-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: kernel:amd64_syscall+0x380
2022-03-03T14:40:11.340464-05:00 <0.7> (id12) /boot/kernel.amd64/kernel: --------------------------------------------------
2022-03-03T14:40:11.340471-05:00 <0.3> (id12) /boot/kernel.amd64/kernel: *** FAILED ASSERTION !domid_vec_empty(new_domids) && !pdm_is_non_ifs_vec(new_domids) ==> new_domids->domids[0] == PDM_ROOT_DOMID @ /b/mnt/src/sys/ifs/pdm/pdm_core.c:660: Invalid membership: [ 1.000000074fc.0001 ]
 

Caused by the following defect: PSCALE-118595

This issue is related to ADS (alternate data streams) workflow. Alternate Data Streams is a data structure used within Windows that stores metadata-types of information about a file, such as comments about a file. It does not exist in FreeBSD (which is the underlying file system of OneFS on Isilon), but OneFS supports it and it is treated as a file within OneFS.
 

How to determine whether a cluster is at risk for this issue?

The occurrence of this defect in the field is rare.

For this race condition to trigger, ALL of the following must be true:

The Cluster is running a OneFS version lower than 9.4.0.0.

The file has more than one parent (hard linked file).

The file must have ADS attached to it.

The file's data resides on cloud storage (no local data on disk).

Rename one of the parents links to another domain (for example, a different snapshot location).

Permanent solution: 
Upgrade to one of these OneFS versions or later which includes the fix:

OneFS 9.4.0.0
PATCH: [9.3.0.5_GA-RUP_2022-03] PSP-1956
PATCH: [9.2.1.10_GA-RUP_2022-03] PSP-2075
PATCH: [9.1.0.14_GA-RUP_2021-12] PSP-1722

A workaround would be tricky for this scenario.  We would have to remove affected files from CloudPools and make them available only locally. It is better to upgrade and fix the problem directly.