I'm not familiar enough with TSM enough to know if what I ask is feasible so let me explain what I'm looking to attempt and maybe one of you that are more expert than I can suggest a solution.
We're currently using TSM 5.5.6 on zOS 1.13 to back up our Windows & *NIX servers. We use an IBM 3584 with 3592 tape and TSM accounts for 30% of the tape activity. We're looking at a Luminex gateway virtual solution to replace our physical tape and allow electronic vaulting at our hot site. If you're not familiar with that product, it looks like 3590 tape to the operating system but is completely virtual and stated to be extremely fast. We will be performing a trial of the Luminex solution and I want to perform benchmarks on TSM's performance on it vs. the current 3592. Obviously, I don't want to endanger my production environment by moving workload to a testbed, but can anyone suggest a method that would allow me to run some "test" functions, backup or copy perhaps, in my production environment and then do the same on the Luminex?
I'm doing the same benchmarking with our full volume backups, DB2 Image copies, and HSM ABARS backups and all are very disposable in that I can run the benchmark, easily redirect it to the Luminex using SMS, and delete the test data afterward and not have any production workload go to the Luminex solution. I'm just not familiar enough with the TSM environment to know if the same type of benchmark can be created but it's very important to measure the performance.
Any advice that can be provided would be most greatly appreciated.
Kind Regards,
Dennis
We're currently using TSM 5.5.6 on zOS 1.13 to back up our Windows & *NIX servers. We use an IBM 3584 with 3592 tape and TSM accounts for 30% of the tape activity. We're looking at a Luminex gateway virtual solution to replace our physical tape and allow electronic vaulting at our hot site. If you're not familiar with that product, it looks like 3590 tape to the operating system but is completely virtual and stated to be extremely fast. We will be performing a trial of the Luminex solution and I want to perform benchmarks on TSM's performance on it vs. the current 3592. Obviously, I don't want to endanger my production environment by moving workload to a testbed, but can anyone suggest a method that would allow me to run some "test" functions, backup or copy perhaps, in my production environment and then do the same on the Luminex?
I'm doing the same benchmarking with our full volume backups, DB2 Image copies, and HSM ABARS backups and all are very disposable in that I can run the benchmark, easily redirect it to the Luminex using SMS, and delete the test data afterward and not have any production workload go to the Luminex solution. I'm just not familiar enough with the TSM environment to know if the same type of benchmark can be created but it's very important to measure the performance.
Any advice that can be provided would be most greatly appreciated.
Kind Regards,
Dennis