This topic has been marked solved and closed to new posts due to inactivity. We hope you'll join the conversation by posting to an open topic or starting a new one.
two days ago after a disk test i got an email from my readynas 314 that disk 1 had to be reallocated and it later added this could mean a disk is failing.
status quo all are working fine. some google search adviced i keep track of these if they allocation errors increases it will be time to replace a disk. so my first question is how do i get to see this count. today i downloaded raidar 6.5 and it gave a number 1688 reallocation errors on disk 1? is it time to replace a new disk then?
i got this readynas 6 years ago in 2014 and i had then bought 4 X 3 TB Seagate Barracuda SATA III 7200rpm. i can find this only in amazon and it says old model. i am assuming my other 3 disks are doing fine. can i buy a different model like seagate iron wolf 3 TB and 7200 rpm ? or even other brands and keep the 3 TB and rpm same. i am wondering the other three will eventually die out and thought get something newer model than getting a seagate barracuda for $77 that way when others fall out i can replace it with newer model of same type.
any suggestions on what i should consider while buying a replacement disk ?
today i downloaded raidar 6.5 and it gave a number 1688 reallocation errors on disk 1? is it time to replace a new disk then?
Definitely it is time to replace it.
Opinions on when to replace a disk vary. Personally I will replace a disk before the reallocated sectors count reaches 50 myself. But everyone would say that ~1700 reallocated sectors is a failed disk.
i got this readynas 6 years ago in 2014 and i had then bought 4 X 3 TB Seagate Barracuda SATA III 7200rpm.
Don't get a new Barracuda - several of those are now SMR drives, and aren't good options for RAID arrays.
If you want to stick with 7200 RPM, then I'd suggest a WD Red Pro or a Seagate Ironwolf Pro. Currently shipping models start at 4 TB, so you would need to go with a bigger size. Note you'll need to upgrade two drives to the larger size to get more space in the volume. The capacity rule is "sum the disks and subtract the smallest).
Other enterprise-class SATA drives would also work (WD Gold to give one example).
This disk is definitely bad! The ReadyNAS have a detection system to ensure a disk is indeed bad before alerting you. You can read more about that here in this link
So, the NAS is already doing the checks for you - i.e. are the errors increasing, etc. If it exceeds the thresholds as described in the above link, you get an alert. At that point it is time to replace the disk.
Please do ensure that you have an up-to-date backup. When disks are getting older, they will be more prone to failure. Replacing the bad disk with a new one will kick in re-sync of the RAID and this is a "strenuous" task for the disks. It is not uncommon to see disks with a long life span, fail during a resync because they were close the end anyway. A disk failure during the sync of your RAID 5, will result in RAID failure and total data loss.
today i downloaded raidar 6.5 and it gave a number 1688 reallocation errors on disk 1? is it time to replace a new disk then?
Definitely it is time to replace it.
Opinions on when to replace a disk vary. Personally I will replace a disk before the reallocated sectors count reaches 50 myself. But everyone would say that ~1700 reallocated sectors is a failed disk.
i got this readynas 6 years ago in 2014 and i had then bought 4 X 3 TB Seagate Barracuda SATA III 7200rpm.
Don't get a new Barracuda - several of those are now SMR drives, and aren't good options for RAID arrays.
If you want to stick with 7200 RPM, then I'd suggest a WD Red Pro or a Seagate Ironwolf Pro. Currently shipping models start at 4 TB, so you would need to go with a bigger size. Note you'll need to upgrade two drives to the larger size to get more space in the volume. The capacity rule is "sum the disks and subtract the smallest).
Other enterprise-class SATA drives would also work (WD Gold to give one example).
Thank you this was informative especially the link with the reallocated sectors. I do a weekly backup and did a round of leftovers and backed them now. Fingers crossed after the new raid syncing.
If you want to stick with 7200 RPM, then I'd suggest a WD Red Pro or a Seagate Ironwolf Pro. Currently shipping models start at 4 TB, so you would need to go with a bigger size. ——————
Thanks again for your reply. One question i have on disk buy : i have three other barracudas that are 3TB @ 7200 RPM cache 64mb if i get a wd or iron wolf that is 4TB and higher cache will that still be okay. If i get the new disk at 5400 isthere any harm different disks spinning at different speeds?
Assuming other three disks are fine if i do end up getting two disks apart from disk 1 which had these reallocated sectors which other disk can i swap/replace the new second disk with?
One question i have on disk buy : i have three other barracudas that are 3TB @ 7200 RPM cache 64mb if i get a wd or iron wolf that is 4TB and higher cache will that still be okay.
Yes. The disk size matters a lot (and as I said SMR disks should be avoided). Cache size doesn't.
One warning here - Netgear does recommend backing up the array before you replace disks. When you replace a disk, the system has to read every sector on the remaining ones to rebuild the array. That does stress the disks (and can uncover issues you didn't know were there). So it is a good idea to update your backup (or make one) first - even if that means you have to purchase USB drive(s) to do that.
@aniraghome wrote: If i get the new disk at 5400 is there any harm different disks spinning at different speeds?
If you mix speeds, then the RAID will slow down (likely somewhere between a pure 7200 rpm array and a pure 5400 rpm array) - but it should still work. I'm not mixing speeds at the moment, but I have done that in the past for some years and had no issues at all. The NAS will give you a warning about mixed speeds if you add a slower disk, but it will let you do it.
I think @Sandshark did run into an issue (spin up time at boot?) where he suspected mixed RPMs were at fault - I don't recall the details or the strength of the evidence. Hopefully he'll join in and clarify that.
FWIW, although your Barracudas seem to have worked out for you, there are quite a few posters who had issues with them (and BackBlaze found them to be unreliable in RAID arrays). Given their age and track record, you might consider schedule replacing them over time, and not wait for them to fail. If you want to go with the less expensive (and cooler) NAS-purposed drives, then you could get one or more WD Red Plus (not a WD Red - they are SMR) or Seagate Ironwolf drives. You'll get the warning, but it should work out ok during the transition period when you are mixing them.
In looking at costs and avoiding smr’s i ended up with two choices. One is wd red plus which i see amazon and most othere doesn’t have stock till jan 3rd.
Dont want to push my luck So planning to go with “Seagate IronWolf 4TB NAS Internal Hard Drive HDD – CMR 3.5 Inch SATA 6Gb/s 5900 RPM 64MB Cache for RAID Network Attached Storage ST4000VNZ008/VN008)“ i plan to buy two this month and two next month just to spread the cost. When i get these new disks can i pop out disk 1&2 and pop these two in and let them sync?
I have a 5tb ext hd which i make 2nd copy of nas data. So from a data perspective i have a backuped them. When you say “while swapping disks netgear suggests backing arrays” how do i back them ?
Dont want to push my luck So planning to go with “Seagate IronWolf 4TB NAS Internal Hard Drive HDD – CMR 3.5 Inch SATA 6Gb/s 5900 RPM 64MB Cache for RAID Network Attached Storage ST4000VNZ008/VN008)“ i plan to buy two this month and two next month just to spread the cost. When i get these new disks can i pop out disk 1&2 and pop these two in and let them sync?
Personally I test the disks first in a Windows PC using vendor tools (Seatools for the Ironwolf). I run the long non-destructive read test, followed by the full erase / write zeros test. You can connect the disks either via SATA or with a USB adapter/dock.
While some might think this is overkill, I have had disks fail out-of-the-box - including one I purchased last month. And I have had a couple of disks in the past that failed one of two tests, but not the other.
After testing I hot-swap the disks (with the NAS running). You do this one disk at a time, and wait until the resync is fully completed before you do the next one. You will see your volume expand by 1 TB when the second disk is synced, and it will expand an additional 1 TB for the third and fourth disks - giving you a 12 TB volume. The NAS reports this in TiB (which is ~10.9). Performance will drop during the resync, and it will take quite a while for each drive.
@aniraghome wrote: I have a 5tb ext hd which i make 2nd copy of nas data. So from a data perspective i have a backuped them. When you say “while swapping disks netgear suggests backing arrays” how do i back them ?
Just make sure your USB copy is up to date - in case something goes wrong.
I think @Sandshark did run into an issue (spin up time at boot?) where he suspected mixed RPMs were at fault - I don't recall the details or the strength of the evidence. Hopefully he'll join in and clarify that.
The problem I had was with OS4.2.x and drive spin-down enabled. Every few months, it appeared that one of the slower RPM drives didn't spin up in time, and was deleted from the array. Sometimes it was then re-detected and a re-sync began and other times I had to force it. That adds a lot of activity to the drives, potentially affecting life and definitely affecting access speed during the re-sync, and also puts you in a non-redundant situation during the re-sync. Disabling drive spin-down fixed the issue. I left the unit powered on continuously, so do not know if I would have run into a similar issue at power-on.
I never re-enabled drive spin-down on that NAS until the drives were replaced, even after updating to OS6. So I do not know if it is more tolerant of that situation. I have some 5400 and 7200RPM drives I've upgraded along the way, and I have an OS6 Pro6 not in use, so maybe it's time to intentionlly create a mixed-drive unit and see how it does. But I've moved all that to storage, so it'll be a bit before I get it going.
in OS6 there is an option Disk-Spin down did unchecking this fix your issue of multiple RPM's? i have this currently checked do i uncheck this ?
@Sandshark wasn't running OS6, but he did say that he'd disabled disk spin down.
I don't think there's a lot of certainty here on whether you need to do this or not, as I ran a mixed RPM RAID array on my pro-6 for some years (also running OS 4.2 at the time) and never saw this problem. I don't see enough data points to generalize.
But if you are concerned, you could uncheck it until all your disks are updated a month or so down the road. You'd use more power, but that's the only downside.
i have only one SATA III to USB cable and will run a short generic and long generic before i load it to NAS. is that good enough ?
Well, as I said earlier, I have had drives that pass the long generic test but fail the erase/write test (and vice versa). So I run them both. What is "good enough" is up to you.
i ran the HDD's through seatools for windows. i did a short,long test but, did not see an erase/write test as an option . is this in a different application ?. long test completed in an hour is that normal ? ofcourse my HDD's are new and with no data.
i am using a docking station to load up these disks for testing will that still test these disks properly as if i connected to my internal SATA port in my desktop?
i have two SATA ports already connected to two HDD's in my desktop hence i went with the docking station option. is there any other s/w i can use in addition or in lieu of seatools?
Clearly i did not do the erase test. But when i did the long generic it did show 100% complete and said passed. These are brand new empty hdd does it still take several hours. Anyway after it passed i swapped them and they resynced and been online past 2 days. Can i turn off readynas take them out and rerun long generic again? What if it still runs 1 hr and says passed. Infact i ran long generic on my old one which i took out of nas it even completed within same duration.
@aniraghome wrote: These are brand new empty hdd does it still take several hours.
The long generic test reads every single sector on the disk - whether that sector is used by the file system or not.
To read every sector on a 4 TB disk in an hour requires a read speed of over 1000 MB/s. Mechanical disks simply aren't that fast - they usually have a max sustained transfer rate of ~200 MB/sec. So it should take at least 5 hours to scan every sector, and I'd expect it to take longer in practice.
@aniraghome wrote: Anyway after it passed i swapped them and they resynced and been online past 2 days. Can i turn off readynas take them out and rerun long generic again?
Not sure why you would need to test them again, if the array is working. But you can power down the NAS and retest them if you like. Don't do the erase test now, since that would require you to set up the NAS again.
It would be good to sort out what is going on with seatools when the other replacement disks arrive.
