I’d like to take advantage of the ability to schedule automatic, incremental backups by using the internal backup capabilities of the EX4. I also want the extra protection of a RAID 5 volume. To do this, I’m trying to select only three of the four installed (2TB WD RED) drives and configure as RAID 5, and then the remaining drive as JBOD. I would then create a share/folder on the single JBOD drive, and use this for internal backup of the critical data (source folders) on the RAID volume. I’ve tried to select individual drives when specifying the RAID volume, but that doesn’t work. I’ve tried specifying only 6TB of the total 8 TB using the slider. The only option I get then is to use the remaining space as a Spanning Volume. Can anyone (WD mods?) confirm that the strategy I’m trying to implement will work, and explain how I can do this. I tried calling customer support and gave up after being told that I need to read page 52 of the guide (which I have done, several times - but it doesn’t say much more than I’ve already described above).
As far as I know you can’t set a dual mode that creates a RAID with a number of drives and the reamining one left out as JBOD with the EX4.
There is an option to create a spare disk when configuring RAID 5. I can not confirm how this works due to the fact that I returned the EX4 and am waiting for the next firmware update instead of spending my time finding and reporting bugs - they used to pay people to do that.
Maybe some one can confirm that by creating a spare disk when creating a RAID 5 volume, that spare disk could be configured as a separate volume. I think there is an option to configure the remaining disk as spanning, but if that option is not selected, I assume that the remaining disk could be used by itself as a separate volume. This configuration would be my choice to enable internal back ups to the spare disk which could be removed and stored off-site.
I tried the RAID5+spare option. I can’t verify 100%, but it seemed to me that this still doesn’t allow the spare to be used for backup. The intent is to have the spare available (and unused) so that if there is a problem with the 3 drives under RAID5, the faulty drive is automatically taken offline and the spare is rebuilt in its place. If there were any data on the spare drive, it would be overwritten, so I doubt it can be used.
But I’m not an expert on the topic.
I suspect that you are probably correct that the spare can not be used in a second volume. I am no expert, either, and it’s unfortunate that WD support also doesn’t seem to know how their own product works. Seems that WD could make a call to India and find out.
If RAID 5 + spare is selected, I assume that the 4th disk will just sit there unused until a disk fails in the array at which point the spare is used to rebuild if autobuild is enabled. In a mission critical application, maybe this would be a required feature. For the average consumer, though, having a disk sit there unused is not a good use of resources and would probably never be used. On the other hand, the desired configuration of a 3 disk array using RAID5 with a 4th disk accessible as a second volume would be a nice configuration for data redundancy and backups. Too bad the design and implementation of this NAS falls short of providing useful working features.
This is a typical implementation of RAID technology going way back. If the drive is in the array unit it is part of the array. As identified you can have 3 of the 4 drives as active members with the 4th drive sitting as a hot spare in the event of a failure of one of the 3 drives. The spare has never been separately usable.
As others have said having 25% of your storage just sitting and doing nothing seems to be a waste. However, if your data is that critical, and you don’t have a suitable backup process, the hot spare makes a lot of sense. If you are paying attention to your unit and the dashboard you will be notified of a drive failure when it occurs and, if it’s convenient, you can quickly purchase a replacement drive and insert it. All part of the “hot” technology. If purchasing a new drive is not convenient you can always keep a spare on the shelf. Of course you now have unused capacity sitting on the shelf. You can’t win!
Either way RAID will rebuild the missing data as soon as the new drive or spare is recognized. If you are doing the 3 drive configuration you can test the process by removing one of the active drives and seeing how the fail over works. It’s best to have a full backup on another device before testing!
ags000 wrote:
I tried the RAID5+spare option. I can’t verify 100%, but it seemed to me that this still doesn’t allow the spare to be used for backup. The intent is to have the spare available (and unused) so that if there is a problem with the 3 drives under RAID5, the faulty drive is automatically taken offline and the spare is rebuilt in its place. If there were any data on the spare drive, it would be overwritten, so I doubt it can be used.
But I’m not an expert on the topic.
You are correct. An hot spare is a drive in standby. When an hdd in an array becomes damaged, the array controller swap the damaged disk by the spare one. This is important because in a raid5 where a disk is missing, all disks and controller are under extra stress due to parity calculation and extra reading. is important in a raid5 to rebuild asap.
Highway wrote:
This is a typical implementation of RAID technology going way back. If the drive is in the array unit it is part of the array. As identified you can have 3 of the 4 drives as active members with the 4th drive sitting as a hot spare in the event of a failure of one of the 3 drives. The spare has never been separately usable.
As others have said having 25% of your storage just sitting and doing nothing seems to be a waste. However, if your data is that critical, and you don’t have a suitable backup process, the hot spare makes a lot of sense. If you are paying attention to your unit and the dashboard you will be notified of a drive failure when it occurs and, if it’s convenient, you can quickly purchase a replacement drive and insert it. All part of the “hot” technology. If purchasing a new drive is not convenient you can always keep a spare on the shelf. Of course you now have unused capacity sitting on the shelf. You can’t win!
Either way RAID will rebuild the missing data as soon as the new drive or spare is recognized. If you are doing the 3 drive configuration you can test the process by removing one of the active drives and seeing how the fail over works. It’s best to have a full backup on another device before testing!
The hot spare is crucial because when an HDD fails, the remaining disks in the array start to work with 2x stress. The array controller not only need to read the data from the remaining disks, but needs to read the parity information so the missing data from the damaged disk can be reconstructed.
Is very well docummented raid5 is the worst raid on earth. In professional forums you can find in I.T. raid5 needs to be avoid, and use raid10 or raid6 instead.
There is lot of scenarios when you lost an hdd from a raid5 and the remaining disks goes under extra stress, you lost a 2nd disk and lost the full array. This is why is good to have an hot spare right away to rebuild asap.
If you lost a drive and don’t have an spare at hand, you request the replacement, take 72 hours… you are hours and days at high risk. At least this is how you need to deal with important data and what you found in professional books about raid5 and hot spare. The risk of max stress rebuild in raid5 is minimized in raid10 because one disk have an exact copy of the other member disk.
Netpro wrote:
The hot spare is crucial because when an HDD fails, the remaining disks in the array start to work with 2x stress. The array controller not only need to read the data from the remaining disks, but needs to read the parity information so the missing data from the damaged disk can be reconstructed.
Is very well docummented raid5 is the worst raid on earth. In professional forums you can find in I.T. raid5 needs to be avoid, and use raid10 or raid6 instead.
There is lot of scenarios when you lost an hdd from a raid5 and the remaining disks goes under extra stress, you lost a 2nd disk and lost the full array. This is why is good to have an hot spare right away to rebuild asap.
If you lost a drive and don’t have an spare at hand, you request the replacement, take 72 hours… you are hours and days at high risk. At least this is how you need to deal with important data and what you found in professional books about raid5 and hot spare. The risk of max stress rebuild in raid5 is minimized in raid10 because one disk have an exact copy of the other member disk.
This is correct. I’m not sure why people use RAID 5 anymore due to this reason. If one drive dies the rest of the array is vulnerable during the rebuild process. There is NO fault tolerance during RAID 5 rebuilds. I think many users take capacity over fault tolerance. RAID 10 allows for a higher level of fault tolerance (1-2 drives) depending which drives fail. If you’re simply looking for capacity go with JBOD because RAID 5 is worthless.
OK, I understand the idea of a RAID 5 array being at risk once a disk fails. But how is this substantially different than when a disk fails in a RAID 1 array? During the volume rebuild when a replacement drive is inserted, if the mirror of the drive fails, then all content on those two drives are lost. Granted, with RAID 5 if any other drive fails all data is lost, but with RAID 1 if the mirror of the failed drive also fails all data on those drives are lost. In large arrays this would mean only a portion of the entire volume is lost, but it may well be the most critical data on the volume. This seems to point back to an important guiding priciple that I didn’t understand until researching more on this topic - RAID is not a replacement for data backup.
