5) Sorry but 10Gb sound like LAN. This question was about SAn, which is probably either 4Gb or 8Gb or 16Gb. If you are able to supply a link to blade specs that would help. See 6) below.
6) Try the FabricOS commands: chassisshow, and slotshow
10b) I think it may be possible that you are confusing bonding with port groups. Bonding is a logical concept, sometimes referred to as trunking, or port-channelling. What I'm asking about it is a physical concept named 'contended port groups'. I think Google will be your friend on this question.
12Q) Just wondering, if I have 3 media server with 7 drives without SSO, 1 media has 2-3 drives. There will be many jobs pending when ever backup policy started. So, this will extend backup window?
12A) If you get the mix of multi-plexing right then it shouldn't extended the backup window. The aim is to meet the backup window. So, assuming your backup clients can deliver the data quickly enough - then there's probably no reason to miss the backup window.
13Q) Additionaly, how to test/get the real drive troughput?
13A) If you have a large capable Unix or Linux server, then GEN_DATA (search ths forum) policies can be useful for testing. If not, then your only method of testing will be with real clients - or to use large files created on the media server - but these won't test backup client read, nor test LAN networking, nor test LAN ingest at the media server.
14Q) Let say, if I configure 4 multiplex (= 4 amounts of troughput in backup job report) so I just sum them all?
14A) It doesn't really work that way. It is possible to view and/or calculate the aggregated bandwidth that a muli-plexed job set did actually experience, but it's deep in the log files. And has been covered elsewhere in this forum.
