Preparing for excam: Questions about ISMV2

Not sure why that is happening. Does this happen for every commend that you post?

glen

No. Not for all comments. But some times my posts are going for a moderator approval and it is taking time to get approved.

Please refer the VNX best practice for performance guide:

https://support.emc.com/docu51389_White-Paper:-VNX-Unified-Best-Practices-for-Performance-VNX-OE-for-Block-5.32-and-File-7.1.pdf?language=en_US

Hi Peter_EMC

This i understand. But if I have let's say 3 disks i can create a RAID 5 with 4+1 and 8+1 what's the difference? For sure there is an error saying please select a mutlipple of 5 or 9 but it is going to work anyway. So i asume 4+1 and 8+1 is the amount of disks in a RAID Group but what are the drawbacks when using 12 Disks using 4+1 or 8+1 is neither a multiple of 5 nor 9.

Any ideas?

Hi Duker,

Raid 5 with a 4+1 (4 data+1 parity) or 8+1 (8 Data+1 parity) is the recommendation from EMC. EMC have done a lot of research on the RAID performance and found that this configuration will give you a better performance. You can create Raid 5 with any number of drives(Min 3 and Max:16 drives), but there will be a performance impact.

Hi Duker,

   In your above post cant it be 8 Data + 1 parity rather 8+1 (7 Data+1 parity)

i am unable to understand this concept of 8+1

8+1 means 8 data and 1 parity

OR

8+1 mean 7 data and 1 parity, if it is right then 7 data and 1 parity are 8 disks including parity one..

bit confusing...:(

If i create a pool it tells me RAid 5 (4+1) number of disks (5 recommended, 10, 15 etc.=multiple of 5) therefore i assume a RAID group persists out of 5 disks (4 data + 1 parity). Well with RAID 5 it's not a dedicated Parity disk moreover it places parity over all 5 disks.

Therefore with a Raid 5 (8+1) it should be 9 disks and parity is spread about all 9 of them therefore you loose the amount of size of the capacity of a disk withing this raid group.

Hi Shr@y,

It was a typo error. it is 8+1 (8 Data +1 Parity). Thanks...

Thank you very much! Finally I've understand the notation and what's behind it!

So from that point of view it's like the following:

notation of RAID means the number uf data and parity disks used (even parity is spread accross, that's the amount of data which is lost).

RAID 5 (4+1) with 100 GByte Disks means= 500 Gbyte raw and 400 Gbyte after parity has been removed.

4+1 and 8+1 defines the number of diskes in a RAID Group which is beeing created which defines the performance/space frame:

4+1 100 Gbyte will lead to the lost of 20 % of space within a Raid 5 8+1 it's the lost of only 12% for parity but it will take longer to rebuild to the amount of disks.

But would a RAID Group with 9 disks not  be more performant than a Group with only 5? And less capacity used for parity? Why is 4+1 recommended then? Security concerning rebuild? What happens to that fact if you create a pool with 30 disks and RAID 5 (4+1) Raid groups? Will there be huge differene between 8+1 for a LUN regarding performance?

Thank you

Hi All,

   What is the basic and a line difference between Block level (  FC, FCoE and Iscsi) ?

Hi Shr@y,

FC -- Data transfer over Fibre Channel network

iSCSI -- iSCSI is an IP based protocol that establishes and manages connections between host and storage over IP (Can use the existing IP network)

FCoE -- FCoE is a protocol that transport FC data over Ethernet network. It uses the Converged Enhanced Ethernet (CEE) link (10GB Ethernet) to send FC frames natively over Ethernet.

Good Day $ree,

       As per my understanding, IP and Ethernet network are one and the same thing, so iSCSI and FCoE must be similar too..

and what is CEE link, Please tell me,

Thanks

Hi Shr@y,

iSCSI -- SCSI packets over IP.

FCoE -- FC data over IP

FCoE provides consolidation of SAN and LAN traffic over a single physical interface infrastructure. If you are using a CNA (Converged network adapter), it will replace both HBAs and NICs in the server and consolidates both the IP and FC traffic.  You can refer IP SAN and FCoE chapter on ISM book for more details.

Take a look at the second document that I attached a couple of post back - the best practices for performance - look in Chapter 5 - Storage System Sizing - this will explain the concept of performance based on the number of disks in a raid group.

glen