I was suspended today. A HDD lost connection yesterday. Did not notice it until today. I re-plug it in and it came back to life. I am in process of retiring the HDD (copying data to replacement HDD), so in a few hours, the drive will be removed from my drive pool.
Audit shows 100% so I am assuming the data is ok. If the drive was dead, node recovery be impossible. I do wonder if I should use redundancy if I ever setup another node because it took a long time to accumulate the data.
An update. So I was able to add a new disk and migrate data so node is now healthy but when I setup the storage pool, I choose “Simple” Layout for the storage. Having issues removing the bad drive. I am using windows server. Using the storage pools option to combine multiple HDDs. I executed the repair virtual disk command, waiting for it to complete and see if I can remove the bad drive then.
First time dealing with the simple layout on windows server storage pools. I do understand there is no redundancy but the HDD came back to life and I managed to migrate 99% of the data from it. There is 2 GB left but I think that is from windows server storage system.
I’ve seen some different opinions on this forum regarding redundancy versus simplicity/low-cost. Personally I run a node on a ZFS file server that I would be using regardless so I benefit from having some resiliency if a drive fails. You’re right that it takes quite a while to accumulate data but only moments to lose it!
I think next time I will setup the pool with redundancy. So it’s looking like that if I can’t remove the bad disk, I will have to do a storj exit procedure, delete the pool and start over.
repair virtual disk is still running. It’s 50% done. Had to reboot machine due to max out memory.
Looks like I got no choice but to do a graceful exit in order to rebuild the storage pool from scratch. Note to self, never use simple layout, use mirror or parity to keep a node online through HDD failures.
So an update, Graceful Exit is still running. I got one satellite left for data to be removed. I been having to re-attach the HDD in Server Manager after reboot the machine after a week or so. System Memory seems to increase probably due to all the file accessing. Something to do with Windows.