NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
OS6 on a legacy device in 2018.
Hello there NAS-owners. Starting this thread instead of me going off topic in the thread about ’restriping’. I just switched from OS4 to OS6 and so far i am happy with the change. The new OS is ...
TeknoJnky wrote:
- no you cannot go from raid 6 to raid 5. You don't necessarily have to factory default, but you would need to delete the data volume and recreate a new raid 5 volume, so the result is the same... you would have to completely backup then restore all data.
Actually if you have an empty slot in the NAS, you can switch to raid-6 without data loss. This would also require that the three largest disks currently in the NAS be the same size.
PerfectOnline wrote:
Are there any ways to verify a directory and save the results in a file so that one can do a quick check if needed? I’m not that educated in this regard but i know of some solution generating .md5 files, i just don’t know enough to put it in practise (best program and stuff like that).
Some years ago I started doing this for selected folders (before I owned a NAS). I still use QuickSFV (though it is old, it still works), creating an "sfv" file in each folder.
You can also use something like QuickPar (also old) or MultiPar for this. (I've used QuickPar, but not MultiPar). These create signature files similar to QuickSFV, but also can create repair files from limited amounts of file corruption. For instance, if the repair files are 10% of the folder size, they can recover from up to 10% of corrupted data blocks in the files. This operates similarly to RAID redundancy btw, but is more sophisticated.
As you say, you can also zip (or rar) the folders - compressed archives also have built-in integrity checks. WinRar does include "recovery records" which are similar to the par2 repair files created by QuickPar. Of course you can't access the data as easily if you do that.
These approaches aren't automated (and if they were, you'd risk updating the signatures when the files became corrupted). So they do take some discipline to use. I don't think it'd be practical to use them on frequently updated folders.
FWIW, btrfs checksums will also detect corrupted files. But it's not easy to access those corrupted files - making repair (or partial recovery) difficult. The bitrot protection is an interesting ReadyNAS feature (which uses the checksums and the RAID partity blocks), but I haven't seen posts here from people saying that it recovered their files. I do have it enabled, but my main strategy for protection from data loss is backups (using snapshots to protect against accidental deletion or modification).
Thanks alot for the reply guise!
I will take a look into this whole thing about .sfv and the likes, it sounds interesting and i do want a feature which i can use if i ever suspect something to be wrong with a file, let’s say a .flac file that sounds strange, it would be sweet to at least know that the file is OK and rule that out, in this example leaving it to a bad rip.
I have enabled bit-rot-protection.
Oh, and a funny thing: since the NAS was worthless to control a single 120mm fan (almost cooked my system) i installed a little thingy that i bought, a voltage regulator so now it is connected to a MOLEX and a single yellow cable goes to the MBU giving me the possibility to see the actual RPM’s, but nothing more than that, the controller is attached on the outside of the NAS so i can control the flow exactly as i want it. 1700rpm’s seems fine, i do think that a system temp if around 50c is too much but i have bought heatsinks and this should not be an issue if my calculations are correct.
Regarding the thing about RAID6-5, i have no empty slots. Would it be possible to remove one disk, shrink the volume and then later add the last disk again or is this too much trouble?
I did the update for the firmware so i am currently running 6.9.5 and current bad things to report are: 0.
Everything seems fine.
I will take a look into this whole thing about .sfv and the likes, it sounds interesting and i do want a feature which i can use if i ever suspect something to be wrong with a file, let’s say a .flac file that sounds strange, it would be sweet to at least know that the file is OK and rule that out, in this example leaving it to a bad rip.
I have enabled bit-rot-protection.
Oh, and a funny thing: since the NAS was worthless to control a single 120mm fan (almost cooked my system) i installed a little thingy that i bought, a voltage regulator so now it is connected to a MOLEX and a single yellow cable goes to the MBU giving me the possibility to see the actual RPM’s, but nothing more than that, the controller is attached on the outside of the NAS so i can control the flow exactly as i want it. 1700rpm’s seems fine, i do think that a system temp if around 50c is too much but i have bought heatsinks and this should not be an issue if my calculations are correct.
Regarding the thing about RAID6-5, i have no empty slots. Would it be possible to remove one disk, shrink the volume and then later add the last disk again or is this too much trouble?
I did the update for the firmware so i am currently running 6.9.5 and current bad things to report are: 0.
Everything seems fine.
You cannot shrink a volume on a ReadyNAS. So if all slots are full, you have to backup, wipe the volume, create a new RAID6 volume, and restore data.
- I see, i am running RAID6 now but after thinking some more about it i’ll keep my current configuration.
I still think it’s strange that i have corrupted data but it must have ocurred when the NAS was busy restoring a drive and me copying stuff from the NAS in the meantime.
If it happened some other time than that it leaves me to doubt the safety of the NAS. I want my data to be uncorrupted.
I remember there being a function under OS4 where you could see if there were any errors during the traffic on the connection but there doesn’t seem to be a function like that anymore when running OS6. I think it were called packages or something like that.
PerfectOnline wrote:
I remember there being a function under OS4 where you could see if there were any errors during the traffic on the connection but there doesn’t seem to be a function like that anymore when running OS6. I think it were called packages or something like that.I still have 4.2.31 running on my pro-6, and I don't know of any such package.
You can download the log zip file and look in network_settings.log.
- On my little Duo it is in: Home/Network/Interfaces/Show errors
I’m almost certain that the same function was present on the RNDU6000 in OS4 but i could be wrong though.
Does network errors always equal corrupted files?
PerfectOnline wrote:
On my little Duo it is in: Home/Network/Interfaces/Show errorsOk. Just download the log zip file and look in network_settings.log That will show you the same stats.
PerfectOnline wrote:
Does network errors always equal corrupted files?No. SMB uses TCP/IP (as do most file sharing protocols). Network errors will result in re-transmissions of the data, so usually the files aren't corrupted. If the connection times out due to the errors, then a file might not be completely sent, and in that case it would be corrupted.
Still, if you are are seeing a lot of network errors you should figure out the cause and fix it.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
