NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Recommended Approach for upgrading drive
I have to replace a drive with some early smartctl errors, looking for advice on how to best proceed. I am going to be upgrading the size and also moving volumes round between a couple of drives tha...
You cannot have two volumes of the same name nor is it easy to change a volume name (see Changing-the-volume-name-on-an-os6-based-nas if you really want to do it). To get what you want, you'll need to use your backup to restore volume contents after you DESTROY and then replace each or an intermediary drive/volume. Since using an intermediary drive entails two copy operations, NAS-to-NAS copy from the backup may not be a lot slower.
To use an intermediary drive:
Add Volume3 in the empty slot.
Copy contents of Volume2 to Volume 3.
DESTROY Volume2.
Add new Volume2.
Copy contents from Volume3 to Volume2.
Delete contents of Volume 3.
Copy contents of Volume` to Volume 3.
DESTROY Volume`.
Add new Volume1.
Copy contents from Volume3 to Volume1.
You could use an external drive as intermediary instead, but it will take more time. Restore from backup is probably at least as fast.
If the drive in slot 1 isn't the primary volume (contains home folders and apps), then there is more to do. See How-to-save-your-apps-when-destroying-your-main-volume-OS6. If that volume was the first drive you put in, it should be primary. Otherwise, one of the others is. This extra step is needed if you use home folders or have apps installed, regardless of the method used to copy the volume contents.
FYI, the SSD with the OS is probably doing less than you think. The OS is in RAID1 on all drives, so having one with the OS only doesn't speed up OS access. It will speed up home folder access and app access if it's the primary drive.
And two cautions: Its best to use different share names when copying to an intermediate drive. Dupklicxate share names can confuse the NAS. Also, if you are using SSH and cp to copy, DO NOT simply "cp" a share. Shares are BTRFS sub-volumes. You must create the share via the GUI before you copy the contents.
I was afraid of that. I will just restore from the backup NAS, that is way more straight forward.
When I pull a drive out (in order to create a new volume of the same name), do I need to do anything else before or after pulling it to actually remove the volume completely, in order to create a new volume of the same name? Or just pull it and its gone? Does it need to be "deleted" in some way through front view or otherwise?
The reason I arranged things the way I did with an SSD for the OS was not for speed per say, it just happened to be a drive I had laying around. That is indeed the drive with the OS. the main reason I did it like that is so that the OS drive is completely separated from my data volumes, I can change my various data volumes at will without affecting my OS installation. So I don't think I should have to move any apps or do anything other then setup the new volumes and restore to them. worst case if the OS drive fails, I'll deal with it, not the end of the world, but again it will be separated from my data volumes.
I realize my configuration is non standard. I am trying to avoid raid5 more than anything. Its possible I will mirror one of the volumes later. Anyway this is how I chose to do it this time. Perhaps it would be ok in the future to have two mirrored volumes, in which case the OS would be on one of those mirrored arrays and the other array would be only a data volume, something like that. But I more or less just decided I don't want to deal with even mirroring and I'm relying on my NAS backup and offsite backup for now.
also for clarity, I am not using X-Raid, using flex-raid... All volumes were created as JBOD.
I have already DESTROYED the other two volumes in preparation to restore to new devices.
However now getting a weird error, I have rebooted it several times to flush everything out, but now when I try to use front view it is prompting me dozens of times and forever for admin login and password, its not remembering it for any length of time, I have tried several different web browsers same result. Any idea what might be suddenly causing that?
I'm wondering if something in the OS was using some space on one of the other volumes I have DESTROY'd. Not sure why it would be, but that's the only thing I changed, destroyed both volumes, there is only one drive in the readynas now, the OS drive...and it SHOULD be functioning fine...but its becoming unwieldy to have to re-enter the admin login and password over and over again.
Also when I tried to then insert one of the new drives and add a flex-raid JBOD volume, its coming back with an error about can't find device or something like that.
So... somehow destroying the two non-OS JBOD volumes seems to have broken my OS functionality.
but its becoming unwieldy to have to re-enter the admin login and password over and over again.
If the password hasn't changed, then this must be a browser issue.
I had IDrive installed in /opt. That's the standard place their installer installs it. When I destroyed these two data volumes, idrive has a deamon running which freaked out and filled up root / in some way. Not sure why, I will look into it later. Anyway, I moved the iDrive stuff out of /opt for now and then the admin login and password work correctly. in the future I will try to see if I can install idrive in /apps which would not have this problem in the future.
However another problem I know have is that somehow when the / filesystem filled up while destroying those two data volumes, when I try to add one of the volumes back using the same name as before, it won't work...complains about can't find something, its not clear..my guess is that when the / filesystem filled up, it blocked the destroy operation on that volume from completing its task, so there is some half-baked config files somewhere now, and I sure hope one of you guys knows how I can clear it out of there so that I can create the volume again with the same name, otherwise I guess I have to factory reset and start all over A-GAIN...which makes my wince..
also since DESTROYING those two non-OS JBOD volumes, now suddenly I see from command line that my / partition is reporting 0 free... which doens't make sense, but that must be why I'm generally having problems, so I'll try to track that down.
So I hope you are the one that can help me figure this out. When I issued DESTROY command on those two data volumes, a running IDrive deamon freaked out and started filling up my / partition really fast for some reason (that I will look more into later). IDrive command line tools is installed by default in /opt. This froze up some functionality with / full.
But the worst part is that I think one of the volume destroy's did not complete its task completely due to the / filesystem filling up halfway through or something, because now when I try to create a new volume with that name, it complains and issues error. I don't know if there is something in the OS configuration somewhere that needs to be completely reset manually to overcome this, or if the actual new drive I'm trying to use was only halfway blessed and needs to be completely wiped in some way to start over or what, but I am hoping you are the one that would know the details of how I can fix this, hopefully not requiring me to factory reset......
But the worst part is that I think one of the volume destroy's did not complete its task completely
What happens if you boot up the system with the original drive in place?
Yea the original drive has already been reformatted and re-volumed and actually restored and is working. I had tried to reboot several times. I had also tried to re-volume this problematic new drive as some other new name which worked fine, but then trying to go back to the first name, complained about couldn't find a device...
But overnight I was restoring the other old drive (that one that used to be called volume2 but is now called volume1...and overnight that drive completely restored everything with rsync and working perfectly with the name volume1. And strangely, this morning I am able to re-volume this new drive as volume2....so somehow the problem seems to have gone away magically.
I don't like it when problems go away magically....that doesn't mean the problem is gone per say... hehe. But at least as of now I appear to have the smart error drive removed, the previous volume2 drive renamed to volume1 and restored, the new drive named as volume2 and in the process of being restored....so with any luck...I'll be good now...
Just as a sidenote, should I have any other concerns about the fact that my / filesystem filled up, presumably by idrive deamon when I destroyed two data drives. Can anything else have gotten corrupted that would give pause to consider doing a factory reset to have completely clean OS? df says my / is 25% full, does that seem like about normal?
I am working with IDrive support now to see if there is a way to install it in someplace other than /opt to avoid this every happening again, their installer doesn't give us the option to specify a path. I could also symlink /opt/IDriveForLinux over to /data/. which would at least work temporarily until I update idrive again probably it would require manual intervention every time. In the past, long time ago, I remember I once had to sym link /var/log over to /data/. because something was logging crazily in there, I don't remember what now.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
