Ensure that your Cisco switch supports EtherChannel (or Link Aggregation Group). It's a common feature on Cisco switches, but it's essential to confirm.

decommissioning the previous datastore if you're satisfied that everything was correctly transferred

If you are utilizing LUNs, you might be able to establish a new one inside the current datastore and migrate data across

Select the method of splitting, build a new datastore with the required capacity, and transfer data from the old datastore to the new one.

Before you do anything, make a backup of your data. To avoid losing any data during the split, the backup should be kept apart from the primary data storage.

Splitting a 190 TB all-flash datastore with 160 TB occupied involves a careful approach to ensure data integrity and minimal disruption. First, identify the virtual machines (VMs) and their storage requirements within the datastore. Prioritize critical VMs and plan for downtime if necessary. If you have vCenter Server, use Storage vMotion to migrate VMs to another datastore with available space. Alternatively, you can add additional storage resources such as new disks or storage arrays to your existing infrastructure. Ensure your backup solution is up-to-date before making any changes. If you don't have extra physical storage, consider utilizing storage features like Storage DRS (Distributed Resource Scheduler) or Storage I/O Control (SIOC) to balance the existing load across multiple datastores within a storage cluster.

this will be a very complex operation, and the steps will vary depending on the infrastructure and requirements. you may try the below:

1. back up your data before making any changes. backup should be separate from the existing data store to prevent data loss during the split
2. choose splitting method and create new datastore with the desired capacity and migrate data from the existing datastore to the new one.
3. you might be able to create new one within the existing datastore and move data in between if you are using LUNs
4. testing after the migration and monitoring
5. decommissioning of old datastore once you are confident that all data has been migrated successfully