12-11-2016 06:10 AM
Dear Experts,
Need an urgent help for master server and media server sizing.
I have an environment in which master server must be on VM and no devices will be presented to it. While Media Server will be handling devices like Appliance(used as an Storage-Server) and Tape Library (12 LTO Tape Drives).
Since, in infra we have VMBackup(arround 120 VM), Oracle DB (More than 100 DB Instance), MS SQL hot(8 Servers) and File-System Backup that has to be done first on Disk and over tape by SLP.
Situation is Master Server must be on VM to achieve high availabilty with VMware Fault Tolerance and Media Server has to offload production DB Server for Full Database Backup. Hence DB Snapshot will be mounted on Media Server and Media Server will run Hot Backup on Disk-OST using RMAN. While archive backup will run from production server in every 4 hours.
Production Site | ||||
Sub-Category | Make | Machine Type | Operating System | Quantity |
Backup Master Server | HP | VM Machine | RHL 6.6 ( Red Hat Linux) | 1 |
Backup Media server | HP | Physical Machine | RHL 6.6 ( Red Hat Linux) | 4 |
Sub-Category | Make | Model | License Capacity | Quantity |
Backup Software | Symantec | Net Backup 7.6 | 240 TB | 1 |
Virtual Library | HP | Store Once 6500 | 288 TB | 1 |
Sub-Category | Make | Model | Drive Model | Quantity |
Physical Library | HP | MSL 6480 ( 12 Tape Drives) | LTO 6 | 1 |
12-12-2016 12:34 AM
Please review the planning and performance guide for suggestion on sizing.
12-12-2016 04:57 AM
Really think that is too many LTO 6 tape drives on one server.
Unless it is a real brute of a machine with extremly good I/O infrastructure, you will not be able to feed all the tape drives.
12-12-2016 11:57 AM - edited 12-12-2016 11:59 AM
Hi Mazhar - unfortunately your question is one of those types of question which is extremely difficult for anyone to answer without *all* of the environmental facts, e.g. solution requirements docs, HLD docs, LLD docs, network topology diagram, application solution docs, database design docs, etc.
Plus even if we did have all of those docs, then you're looking at about one to two man weeks of detailed design work, and close contact with solution designers from all areas.
To me, it would seem that someone has already done the scoping and sizing, because you appear to have a loosely defined kit list. Maybe it might be worth going back to the team that generated the kit list to find out how they determined what the kit list should be. i.e. someone must be able to demonstrate why the kit meets the design which meets the requirements which meets the budget.
If someone has picked that kit list out of thin air, then... IMO, you'll be quite lucky to get it to work. It may well work very easily, and you find that you've massively overspent needlessly on too much backup infrastructure.
Riaan's post will help you. The document that Riaan has pointed you towards will not give you a completely worked process for sizing, nor validate your design. The document that you have been pointed towards is meant to help you begin thinking in the right way about what matters in backup planning, and especially about what matters in NetBackup planning.
HTH. Good luck.