Skip to main content
Solved

Cisco UCS RAID Attributes for DDB/Index


dude
Byte
Forum|alt.badge.img+15

Hi All,

I`m deploying new Media Agents and configuring as RAID 1 for DDB and Index on SSB Drives. My question is related to the best attributes for a RAID that will host DDB / Index. Is this defined by Commvault? Thanks

 

Best answer by Bdas

Hi,

Though the Strip size attribute is not documented . From deployment experience,  64K  will give you a 128K stripe size. This is small enough as to not waste a lot of space, yet large enough where performance with large files should be satisfactory. With regards to DDB volume , 64K should be good enough as the large files is less of a condition with DDB structure  . 

 

Thanks & Regards

View original
Did this answer your question?

3 replies

Jon Vengust
Vaulter
Forum|alt.badge.img+5

Hi, thanks for raising this matter.

 

Commvault does not define specific write policies for DDB volumes. However, if your RAID controller has a battery, then feel free to use “Write Back Good BBU”. As you may be aware, write back improves performance at the cost of being vulnerable to a power outage or similar disaster as the writes are cached first whereas write through is an immediate write to the disk.

 

In either case, it is personal preference.

 


dude
Byte
Forum|alt.badge.img+15
  • Author
  • Byte
  • 320 replies
  • April 2, 2025

How about strip size?

 


Forum|alt.badge.img+3
  • Vaulter
  • 8 replies
  • Answer
  • April 2, 2025

Hi,

Though the Strip size attribute is not documented . From deployment experience,  64K  will give you a 128K stripe size. This is small enough as to not waste a lot of space, yet large enough where performance with large files should be satisfactory. With regards to DDB volume , 64K should be good enough as the large files is less of a condition with DDB structure  . 

 

Thanks & Regards


Reply


Cookie policy

We use cookies to enhance and personalize your experience. If you accept you agree to our full cookie policy. Learn more about our cookies.

 
Cookie settings