Hi arsalankhan,
1. What is the impact of changing the Raid Group size? (free space, performance, restriping, etc..)
No immideate impact. Altering the RG size will only change how many disks can be added to the current RGs or newly created RGs later on. Also setting the RG size to a size lower than the number of devices present in the current RGs will not remove any devices. Those RGs will stay as they are. Only newly added RGs will be smaller. Max RG size is dependand on the RAID level and differs between RAID4, RAID-DP and RAID-TEC. The RAID level can be changed on the fly. By increasing the level from RAID4 to RAID-DP or RAID-DP to RAID-TEC, a reconstruct will kick-in to build the additional parity drive. By decreasing the RAID level, the respective parity drive will just be removed from the RG and put back into the spare pool.
2. Is it Doable on production 7-mode as well as cDOT systems ?
Yes.
3. Right after growing the aggregate by adding additional Raid Group, Will there be any data balancing or the new writes goes to the free Raid Group/s ?
There is no automatic rebalancing. All new incoming writes will be striped across all disks immediately, meaning old and new RGs. Once the old RGs become full, writes will only go to lesser filled RGs, possibly impacting performance. So when RGs are added as long as there is ample space available on the old RGs, the system will rebalance by new writes over time. If the RGs have been filled already (e.g. above 80%) a re-allocation is highly suggested.
4. Any best practices guide to follow here ?
Re-allocate volumes on HDD-based systems when aggregate was filled >80% while disks and/or new RGs were added. No need to do this with SSDs as they provide enough performance.
regards, Niels
---
If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO or both.
