Effect of turning off LVM bad block relocation on a RAID 0+1 XP1024 array

[David Fu_2]

Hi,

This may be a repeat question but I searched high and low on ITRC with zilch results.
I was wondering if it was necessary to use bad block relocation on LVM on a RAID 0+1 array?

Here it goes:
Has anyone ever implemented lvm with bad block relocation turned off (-r N)?

What is the resulting effect of bad blocks appearing on a RAID1 volume and how would that affect the data integrity? Would it suffice to let the RAID hardware handle it?

What's your recommendation on it for hardware RAID 0+1 volumes? Leave it on or ok to turn it off?

Would appreciate your insight and answers.

A Happy New Year to you all!

[Thierry Poels_1 · Accepted solution]

Hi,

for intelligent disk arrays like the XP, EVA, EMC², ... bad block relocation should be turned OFF. Bad blocks will be handled by the hardware, no need handle this on LVM level.

regards,
Thierr

[David Fu_2]

Hi Thierr,

Thanks for the info!

I don't supppose you would know of any tech support doc that has more on this for the XP?
With VA and EMC, I found serveral links to either business support note or technical knowledge docs.

The XP was casually mentioned in certain places but without solid documentation backing. ;P

Happy New Year!

[Tom Danzig]

Bad block relocation should be turned off for the XP disk arrarys. This comes straight form top notch HP-CE when our XP512 was set up.

I also remember reading this as well in XP512 setup and configuration documentation. I don't currently have these doc's available to confirm though.