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.
Got it thanks! Sounds like I will be sticking with xraid 1, but for the sake of educating me, what would other options be with a two-bay NAS...
One option is to use flexraid/jbod to create two volumes (one for each disk). You have no RAID redundancy, and you can't expand either volume. But it does maximimize the space. I use this mode on an old duo v1 to get two 2-TB volumes. If one disk fails, you lose the data on it, but the data on the other disk will remain intact.
It is also possible to create a single volume using RAID-0 (which spans both disks). This also has no redundancy or expansion. If either disk fails, you lose all the data - a huge disadvantage, so I don't recommend that mode.
since stopping the readynas, I reseated both drives just for what the heck, I don't know if that is relevant or not, and restarted, it resynced them and now I have raid1 with both drives seems to be working fine. Anyone have any idea why one of them might have gone to "black" color code and degraded and now working fine? Where can I find some low level logs taht might give some hints, if any, about the drive having problems?
I don't have a PC in order to run typical seagate diagnostic software.
Well that's interesting. I've been using these two drives without any problems whatsover for years. its strange to me that there were no errors in the general log other then "its degraded now". No error messages leading up to that...just suddenly went degraded, and now working fine...
here is what I find in the smart_history.log you mentioned:
Well that's interesting. I've been using these two drives without any problems whatsover for years.
You might find this interesting: https://www.backblaze.com/blog/3tb-hard-drive-failure/ AFAIK this is the only drive model that BackBlaze removed from service before they failed. I guess 32% failure rates were just too high.
WDC Red and Seagate Ironwolf are much better drive choices for RAID.
here is what I find in the smart_history.log you mention
No reallocated or pending sectors, which is good.
The command timeouts on three of the four drives is troubling. If a drive doesn't respond to an ATA command it certainly could be declared dead. One aspect here is that command timeouts can have multiple causes (disks, SATA backplane, and in some rare cases driver problems).
If the drive is declared dead with pending writes, then rebooting could result in a RAID array that is out of sync - which can cause file corruption, or even prevent the array from being mounted.
1 - I only have two drives. I don't know why the smart log has 4 lines. I have two drives in a raid 1 config
2 - I can appreciate the bad news about my drive models, but replacing them is not an option for me at this time unless they truly have gone bad.
3 - Any other advice for how I can determine if one or both drives have actually gone bad or why the readynas would have degraded the system? Especially considering that as of now they are no longer degraded.
1 - I only have two drives. I don't know why the smart log has 4 lines. I have two drives in a raid 1 config
That's my mistake - It's because you are looking in smart_history.log It's showing changes over time. The last line for W1F0B60F is the most recent status, with the command timeouts first increasing from 0 to 2 on 10:11:31 1 January 2017, and then from 2 to 4 a couple of minutes later. That's likely the one that went to "black" and resynced - can you confirm that.
3 - Any other advice for how I can determine if one or both drives have actually gone bad or why the readynas would have degraded the system? Especially considering that as of now they are no longer degraded.
Are you using the volume maintenance schedule? If not you should consider scheduling regular disk tests on the volumes. I run them quarterly, but I'd suggest that you run them more aggressively just in case there's an issue - perhaps monthly. The other maintenance options are worthwhile also, but I'd personally stick to quarterly on those.
when you say "confirm it", what do you want me to do to confirm it? All I know about is the smart log which I posted earlier.
I do have disk test and scrub running once a month already actually... Interestingly, its on the 1st, which is the day of the timeouts...not sure if that's related.
when you say "confirm it", what do you want me to do to confirm it? All I know about is the smart log which I posted earlier.
I mean check to see that the drive with serial number W1F0B60F is the one that turned black and then resynced. You should be able to see drive serial number for each slot by clicking on the volumes settings wheel, click on settings and then click on disks.
I do have disk test and scrub running once a month already actually... Interestingly, its on the 1st, which is the day of the timeouts...not sure if that's related.
It could be related.
I suggest spacing out the disk test and the scrub, and not scheduling them in parallel.
yes that is one and yes they are doing those tasks on different days of month
Then you should get good notice if the problem gets worse. Maybe monitor the smart_history monthly too, and see if more command timeouts are logged (and if they are all aligned with maintenace tasks).
ok a year later....getting the uncorrectable error messages again. if I rsync the drive, the messages go away for a while, but return after a week or two. I'm gonna buy another drive. Question is, do I need to match the same model or can I get a different one without any issues? Currently its the ST3000DM001, which as noted earlier is konwn for problems, so I'm kind of keen to replace it with something better, but not if its going to cause any problems with the other existing ST3000DM001. Also, can I put in a larger drive and then later on come back and upgrade the drive that is working ok to a larger drive?
Using X-Raid in Raid1 config (2 drive bay). so... question is...
1 - Should I just buy another DM001 and be done with it, they are $80 right now for this older model, but it matches.
2 - If not, could I put any other drive model or should I limit which models I might put in there for the defective drive and if so, how would I determine what would be an acceptable drive that is perhaps more resiliant then this model?
3 - If I were to get a larger drive to replace the defective one, say 4TB, I realize that initially it would sync with the first one and I'd have only 3TB of storage, but later on I could replace the good 3TB drive with a larger 4TB one. Then will it resync up to 4TB or would I need to back everything up and reformat the array from scratch in order to get 4TB?
I'd replace the DM with a WDC Red or a Seagate Ironwolf. Both work well in ReadyNAS.
You can use a larger drive if you like, as long as you are aware that the volume won't expand until you replace the second drive with the larger size too.
no issues mixing and matching my DM drive with one of those?
I have been using Xraid (raid 1 mirroring), is that the mode that makes most sense for me continue using if I should opt to get a larger drive, now and won't upgrade the first drive until later?
I already understood that putting a larger drive in now will not expand my volume because drive1 will still be the smaller size, but if and when I upgrade that one to a larger one also, will the xraid automatically increase the size of the volume at that time? In other words, right now I pull out the defective drive and replace it with larger 4TB drive. Resync will automatically occur and I will have 3TB volume based on the smaller drive. Later if I pull out the small drive and replace it with a larger one, readynas will resync again...will it end up automatically with a 4TB volume at that point in time?
I already understood that putting a larger drive in now will not expand my volume because drive1 will still be the smaller size, but if and when I upgrade that one to a larger one also, will the xraid automatically increase the size of the volume at that time?
Yes, it will expand when the other DM is upgraded.
I have been using Xraid (raid 1 mirroring), is that the mode that makes most sense for me continue using if I should opt to get a larger drive, now and won't upgrade the first drive until later?
I think staying with XRAID makes sense. Any other option will require you to destroy the existing volume, create a new one, and restore the data from backup. And the only way to get more space is to give up the RAID redundancy.
Got it thanks! Sounds like I will be sticking with xraid 1, but for the sake of educating me, what would other options be with a two-bay NAS, hypothetically if I did want to reformat from scratch? I'm presuming that in a two drive system, I can't really get any bigger volume size, so I didn't perceive any advantage to using a different raid type... I guess maybe if I got a bigger drive, I could use a different type of raid and use the smaller drive for parity or whatever...maybe? I don't plan to do that anyway, but just asking for educational purposes.
I think staying with XRAID makes sense. Any other option will require you to destroy the existing volume, create a new one, and restore the data from backup. And the only way to get more space is to give up the RAID redundancy.
Got it thanks! Sounds like I will be sticking with xraid 1, but for the sake of educating me, what would other options be with a two-bay NAS...
One option is to use flexraid/jbod to create two volumes (one for each disk). You have no RAID redundancy, and you can't expand either volume. But it does maximimize the space. I use this mode on an old duo v1 to get two 2-TB volumes. If one disk fails, you lose the data on it, but the data on the other disk will remain intact.
It is also possible to create a single volume using RAID-0 (which spans both disks). This also has no redundancy or expansion. If either disk fails, you lose all the data - a huge disadvantage, so I don't recommend that mode.
