Difference between Hostbased , Array based and Switched based Splitters?

Hello,

Greeting the day. Please see the below mentioned updates:

Array-based Splitter:

A splitter that is installed on an EMC storage array, and performs the data splitting function. This feature is only available for EMC CLARiiON CX3 and  CX4 storage arrays. In order to run this splitter on the CLARiiON array, the CLARiiON splitter bundle and CLARiiON splitter enabler must first be installed.

Although the two storage processors of a CLARiiON splitter are listed as separate entities (CLARiiON Splitter 1-A and CLARiiON Splitter 1-B), they are managed as a single entity. This means that when you manually add or remove a splitter, the second storage processor instance is automatically added or removed.

Fabric-based splitter:

An intelligent fabric switch that performs the splitting function. To date, this feature is available for Cisco SANTap and Brocade switches using the following modules:
          • In Brocade – AP-7600B and PB-48K-AP4-18
          • In SANTap SAN-OS and NX-OS – SSM, MSM 18/4 or 9222i native

             Brocade splitter :  An intelligent fabric splitter that is installed on the blade processor of a Brocadebased Connectrix switch or blade, and performs the data splitting function. This                                               splitter is currently available for the following Connectrix devices:
                                                  • AP-7600B switch
                                                  • PB-48K-AP4-18 blade

             SANTap splitter:  A splitter that uses the SANTap services to perform the data splitting function. To date, this feature is available for the following SANTap SAN-OS and NX-OS                                               modules:
                                                   • SSM
                                                   • MSM 18/4
                                                   • 9222i native

Host-based splitter:

RecoverPoint software that is installed and runs on supported hosts that mount storage volumes that are replicated by RecoverPoint. The host-based splitter intercepts or “splits” application writes so that they are sent not only to their normally designated storage volumes, but also to the local RPA. To date, this feature is available for the following operating systems:
     • Windows
     • Solaris
     • AIX

The RecoverPoint host-based splitter consists of two main components, a user space application, which is frequently referred to as the actual “KDriver” and a kernel driver, which is frequently referred to as the “Splitter”. In general, the “KDriver” is responsible for the control path and the “Splitter” is responsible for the I/O path.

Does it make sense? Thank you.

Cheers!!!

Suresh Thirumalaisamy

Technical Support Engineer

Worldwide Technical Support - Virtualization Technologies Support Group

Office Hours: Tue-Sat, 21:00 to 04:00 hrs EST

Phone            : (US Tollfree)1-800-782-4362 || Extn 25083

Email               : Suresh_Thirumalaisamy@emc.com