Wondering if it is even possbile to upgrade from 1TB drives in there IQ36NL nodes to 2TB drives? First off I would think this is technically possible but the question is, would we ever want to do it as i would be very messy and take a long time.
2 scenarios that I could see working listed in the order of ease of implementation from a conceptual standpoint. Performance is not a issue at this time on this cluster so creating hotspots although valid is not really a issue here.
- SmartFail one node out(providing they have enough capacity on the other three and replacing the drives in it and adding it back to the cluster. Rinse and repeat for the rest of the nodes.
- What I don’t know about this is will the system even let this happen as the new node will not match the old ones in hardware spec.
- SmartFail a series of drives across all the nodes. Obviously do this 1 at a time and let the system restripe after each one. Than add the new 2TB drives to each of those slots. Guessing this would create a “new pool” which you could than selectively utilize smart pools to move data to it. Than rinse and repeat until you have all the drives replaced.
- This could take a ton of time.
I am just throwing this out there to see if anyone has even attempted this for my reference as the person I am working with told me that he was told that he could do this when the system was put in.
Solved! Go to Solution.
Re: Drive upgrade in Existing Nodes
Have you considered leasing some 72NL nodes and using a (temporary) SmartPools license?
1 := 1 TB node, 36NL
2 := 2 TB upgraded node, 72NL
L := leased node, 72NL
Build two pools and use SmartPools to migrate:
111111+LL start, single old pool
11111+2LL 2LL: new pool
111+222LL 111: must be finally migrated at once, because two nodes don't make a pool
Depending on the used capacity on the current cluster,
zero to three nodes would be required to be leased.
The IQNL-Series nodes cannot be upgraded to larger disk capacities.
With newer generation nodes (S200/X200/X400/NL400), a field upgrade is possible, however, it requires a Support engagement to update the serial-number-specific manifest file on each node to allow it to recognize its new configuration.