Greetings.
Could somebody explain if it's a good practice (is it worth at least?) to do gjournal on top of a gstrip array, or further, on top of a raid 10 (gmirror+gstripe)?
Of course, considering the following partition scheme (in order to boot up the system):
And at last, what are the considerations to do gjournal on top of raw disks/stripes/mirrors (examples: ad10.journal or gs0.journal or gm0.journal), slices and partitions? Is it safe to journal an entire disk/stripe/mirror or preferable to do that only on partitions?
The main idea behind it, is to safe keep data integrity and skip fsck at all cost. (considering a big databases integration point)
That's all. (I'll keep the questions about GPT plus all that, just for curiosity, for later on
)
For everyone who spends time to clarify these confusing questions, know that it will be very appreciated.
So, thanks in advance.
Could somebody explain if it's a good practice (is it worth at least?) to do gjournal on top of a gstrip array, or further, on top of a raid 10 (gmirror+gstripe)?
Of course, considering the following partition scheme (in order to boot up the system):
Code:
(gmirror)
/boot
(gstripe) or (gmirror+gstripe)
/
swap
/tmp
/usr
/var
/homeThe main idea behind it, is to safe keep data integrity and skip fsck at all cost. (considering a big databases integration point)
That's all. (I'll keep the questions about GPT plus all that, just for curiosity, for later on
)For everyone who spends time to clarify these confusing questions, know that it will be very appreciated.
So, thanks in advance.