cancel
Showing results for 
Search instead for 
Did you mean: 

What exactly means "stripe across" in the New Volume Wizard?

unex
Level 3
Dear all, as you can read in the topic, I want to know what exactly is the meaning of the "stripe across"-function in the New Volume Wizard or when do I need it? There are the options "Port", "Target", "Enclosure" and "Channel". Port, Target and Channel are shown in the Properties of the Harddisk under "Device Type" and the Eclosure will be found in the Navigation-Tree. But how can I exactly use this feature and what for? Performance? Stability? RAID-Controll over hardware instead of software? Would be nice, if you can give me some answers. Best regards, unex Ah, btw: I'm running SF 5.1SP1 on two Windows Server 2003 Enterprise R2 Cluster Nodes with mostly DAS DiskArrays [SCSI] from Infortrend. They are all RAID5, but you see only the full builded volumes as Harddisks, because the Infortrends have their own RAID-Controller onboard!!! Additionally there are RAID5-Volumes from a SAN with EMC-CX10-Hardware, so I have different Enclosure-Entries in the Navigation-Tree.
1 ACCEPTED SOLUTION

Accepted Solutions

Marianne
Level 6
Partner    VIP    Accredited Certified

These are generic storage terms and are therefore not explained in SF documentation. If you go to Windows Device Manager, select a disk and view Properties, you will see similar addressing for each device.
Port normally refers to the path
Target - scsi target assigned to the entire array
If I'm not mistaken, 'channel ' is same as bus.

The idea with 'stripe across' is to avoid striping within the same array, as striping is already done at hardware level.

Have a look at this Device Glossary: http://technet.microsoft.com/en-us/library/cc770791%28WS.10%29.aspx


View solution in original post

5 REPLIES 5

David_Veber
Level 4
Employee Accredited
Stripe across will create raid-0 data stripes across the selected disks.

So you create a volume normally it creates a concatenated volume that spans across multiple luns but using a contiguous adress space on the disk. If you create a striped volume it will span the disks in a striped fashion. Say we are creating a striped volume on 3 disks the following will  happen. One write to the first disks the next write to the second disks the next right to the third disk then the 4 right to the 1st disk the 5th write to the 2nd disk and the 6 right to the 3rd disk and it continues like that.

More information can be found in the admin guide for SF. Here is a link to the SFW Admin guide for 5.1. Just do search for stripe.
http://sfdoccentral.symantec.com/sf/5.1/windows/pdf/SFW_Admin.pdf




unex
Level 3
Hello David, 

thanks for the link and your answer, but I know what RAID0 (aka "Stripped") do with data on the disks. 
My question relates to the special possibility by building a stripe-set and using additionally the option "stripe across", where I can choose Port, Channel, etc.! 
The manual is like the help - only the info that I have the possibility to "stripe across: Port, Channel,..."; great. I want to know the effect on this if I choose it additionally with e.g. Cannel. 

Is there a better way to contact a technician from Symantec or someone who knows about this than over this forum? 

Marianne
Level 6
Partner    VIP    Accredited Certified

These are generic storage terms and are therefore not explained in SF documentation. If you go to Windows Device Manager, select a disk and view Properties, you will see similar addressing for each device.
Port normally refers to the path
Target - scsi target assigned to the entire array
If I'm not mistaken, 'channel ' is same as bus.

The idea with 'stripe across' is to avoid striping within the same array, as striping is already done at hardware level.

Have a look at this Device Glossary: http://technet.microsoft.com/en-us/library/cc770791%28WS.10%29.aspx


unex
Level 3
Hello Marianne, 

thanks for your answer. 
So it goes a little bit in the direction of the "private group protection" if I get it right. 

Marianne
Level 6
Partner    VIP    Accredited Certified

Private disk group protection is actually something else... This feature is meant to provide protection on disks/luns in a diskgroup that is visible/accessible by more than one server.
Extract from Admin Guide:
Private dynamic disk group protection uses hardware locking techniques to protect secondary dynamic disk groups located on shared storage from access by other hosts connected to the shared storage pool. The hardware locking is implemented by using a SCSI reservation thread to maintain a current reservation for each disk in a protected group.