We had a switch stack issue and our 3 SDS node rebooted at same time.
The system came back with no degraded area but making a huge rebuild (55 TB, almost all the data we haver over it).
There were some Path and Original-Path scrambles, I did the update_device_original_path in only one device but we didn't have any change in rebuild process.
Right now I'm limiting the rebuild bandwidth to be able to access decently some data.
The very weird thing is that there's no degraded area.
I tried to disable the Rebuild/Rebalance in Backend at Storage Pools backend but no difference.
Any help/hint would be very appreciated.
Solved! Go to Solution.
G'day Schotten, If the SDS nodes are in the same fault set the only way to ensure data integrity is a complete rebuild. If they are not in the same fault set, the problem should be investigated so please run get_info from the master MDM and each of the SDS nodes, open an SR and attach the get_infos to the SR.
Thanks for the reply.
The SDS were in the same Fault Set. We just waited for the rebuild (took 40 hours) so things went back to normal.