wafl.watch.cache.limit.near: Size of "change notify" requests in inode cache pool: , size limit

[ornate ferry]

Getting on an AFF system

Message: wafl.watch.cache.limit.near: Size of "change notify" requests in inode cache pool: xxxxxxx, approaching the size limit: 104857600
Description: This message occurs when WAFL(R) "change notify" requests approach the size limit of the inode cache pool. This is caused by an excessive number of "change notify" requests being generated by the Server Message Block(SMB) server. If the maximum limit is reached and the limit is not increased, significant system performance degradation could occur.
Corrective Action: Contact Contact NetApp technical support. for assistance.

The system is currently in data migration phase using robocopy from a 3rd party vendor storage to Netapp AFF cifs volumes; volumes are 20tb flexgroup vols ( CIFS/SMB shares) language is utf8mb4 and using -is-search-short-names-enabled true on vserver cifs options level .

Volumes are replicated also via SVM-DR to another system .

Ontap is 9.11.1p8 ; the system has 24x3.8 TB SAS SSD.

What is actually the issue here the inode cache pool ( how do i increase that ? )?

[slow jackal]

The issue is that some client(s) use change notifications (aka directory/file watches) quite liberally, and threaten to overwhelm the memory set aside for those watches in ONTAP. There is no functional impact if that memory runs full, as the client software usually falls back to using polling (i.e. regularly checking files/directory to see if anything has changed). If this happens on a large scale though, it might impact performance.
There is currently no way to see what clients have those notifications open, and how many. However, you can increase the memory limit from 100mb to 200 or 300 mb, depending on the system (and only with guidance from NetApp support). That will, however, almost certainly only push the problem further away, not get rid of it