zpool status and smart values

[Sander Jansen]

hey, i got a message from freenas that one disk is possible getting errors... i have got some zpool information and smart values but at this point im a bit at a loss if i should replace disks and if so, what disks.

it seems i need to replace disk ada9 but do i also need to replace more disks soon? im going to the store to get some more but i need to know how much disks i should get...

(see attachments for the outputs)

at the moment im doing a "zpool clear" because the system was running via a VM for a while before i put it back onto a full hardware environment so i thought maybe something went wrong when it was in the VM state. but after the clear the errors still persisted...

Also is there a way to tell freenas that i dont need an e-mail message every minute that the pool is degraded?

 

[Bidule0hm]

You've drives who have seen 50 °C temps, it's far too high, they shouldn't be allowed to go over 40 °C if you want to keep them in good shape.

ada2 is a WD green with a very high LLC value (500k+), I guess the idle timer hasn't been changed from the factory default, you should read about WDIDLE and do something about that.

Some of the drives have a few CRC errors, usually caused by bad contacts on the SATA connectors or power glitches, nothing to worry too much.

I don't see any drive failing hard right now. Can you post the output of the command glabel status so we can map the GPTID to the device name please?

 

[Sander Jansen]

I decided to replace one drive for now with a 6tb i have... the rest is the same (ada9 was replaced) As the disk indeed still is good, i will place it back in after the rebuild. it was a bit stupid to replace it like this as im planning on slowly (because of costs) creating a new raid6 setup to replace this one but that is going to take a few months...

My initial thought was to start replacing disks with 6tb ones and then kick out the 2tb... but then i came to the realization that i cant do that because of the space i have on the 6tb disks is only 2tb as long as i have 2tb drives in the setup... little mistake there... what i want to do now is just save up disks and create a new raid6 setup and then copy everything over and delete the initial setup... or is there a way to replace the drives one by one?... i was planning on getting 4x 6tb drives instead of the 12 2tb drives i have now... but to do it replacing one by one i would need 12 6tb drives dont i before i can actually increase the size used to 6tb per drive... or is there a better way to do this without getting 12 new drives?

gptid/151b41fb-b885-11e5-b7e7-d050995afa5e N/A ada0p2
gptid/1106f45a-b885-11e5-b7e7-d050995afa5e N/A ada1p2
gptid/13258dfd-b885-11e5-b7e7-d050995afa5e N/A ada2p2
gptid/10280609-b885-11e5-b7e7-d050995afa5e N/A ada3p2
gptid/13fd7b24-b885-11e5-b7e7-d050995afa5e N/A ada4p2
gptid/121b2319-b885-11e5-b7e7-d050995afa5e N/A ada5p2
gptid/0b006576-b885-11e5-b7e7-d050995afa5e N/A ada6p2
gptid/0c9404a3-b885-11e5-b7e7-d050995afa5e N/A ada7p2
gptid/0bc9a539-b885-11e5-b7e7-d050995afa5e N/A ada8p2
gptid/e36951d1-e3d0-11e6-8703-d050995afa5e N/A ada9p2
gptid/0e8518cc-b885-11e5-b7e7-d050995afa5e N/A ada10p2
gptid/0f5a30e4-b885-11e5-b7e7-d050995afa5e N/A ada11p2
gptid/3ffe3acc-e25a-11e6-ad1a-d050995afa5e N/A da0p1
gptid/9aaf3b94-e3dd-11e6-9690-d050995afa5e N/A da1p2

 

[GBillR]

hey, i got a message from freenas that one disk is possible getting errors... i have got some zpool information and smart values but at this point im a bit at a loss if i should replace disks and if so, what disks.

it seems i need to replace disk ada9 but do i also need to replace more disks soon? im going to the store to get some more but i need to know how much disks i should get...

(see attachments for the outputs)

at the moment im doing a "zpool clear" because the system was running via a VM for a while before i put it back onto a full hardware environment so i thought maybe something went wrong when it was in the VM state. but after the clear the errors still persisted...

Also is there a way to tell freenas that i dont need an e-mail message every minute that the pool is degraded?

That is quite the cluster of drives you've got there... those AV-GP drives are not really the best for data storage...
That said, those errors on ada9 don't look to be disk failures. You should run a long test and maybe even a badblock test on it, but I suspect @Bidule0hm is on the money with it being a cable, power, or other issue.

You should read here and here.

And it doesn't look like you are performing scrubs of your pool... with THAT pool, you are taking a pretty big risk in my opinion if you are not performing scrubs.

And I would then wonder if you are performing and SMART long tests as well.... are you?

 

[nojohnny101]

you are correct in your thinking but you are using some wrong terminology. there is no raid6 in freenas. if you want to increase your storage you can either create a new vdev with newly purchased 6tb drives and stripe it to your existing pool or you can create a new pool and create a new vdev in that new pool with any number of 6tb drives you want. or as you said, you can replace all 12 disks with 6tb ones and then your storage will increase.

you are correct to think that storage will not increase until you replace all 12 disks with higher capacity ones. freenas always uses the lowest capacity disk to determine storage size of the whole vdev.

 

[Sander Jansen]

And it doesn't look like you are performing scrubs of your pool... with THAT pool, you are taking a pretty big risk in my opinion if you are not performing scrubs.

And I would then wonder if you are performing and SMART long tests as well.... are you?

Yes im doing both, thats why i started wondering about ada9 because it has reed and write errors... at least that is what the scrub came out with...(see zpool information in attachments)

I needed the space and ive been adding disks ... that is why i have so much disks now... and yes im understanding the risk... that is why im starting to replace with bigger ones... well i have to remove that disk again when its done rebuilding because i need to create a new dataset so yeey for rebuilding twice... well you have to learn at some point...thy say :p

while replacing ada9 i indeed found a half loose cable so that probably enplanes that...

...using some wrong terminology. there is no raid6 in freenas

I guess that is my mistake... i named the "tank" that name... because i came from a hardware raid6... so i sort of kept the name...

 

[Bidule0hm]

I decided to replace one drive for now with a 6tb i have... the rest is the same (ada9 was replaced) As the disk indeed still is good, i will place it back in after the rebuild. it was a bit stupid to replace it like this as im planning on slowly (because of costs) creating a new raid6 setup to replace this one but that is going to take a few months...

My initial thought was to start replacing disks with 6tb ones and then kick out the 2tb... but then i came to the realization that i cant do that because of the space i have on the 6tb disks is only 2tb as long as i have 2tb drives in the setup... little mistake there... what i want to do now is just save up disks and create a new raid6 setup and then copy everything over and delete the initial setup... or is there a way to replace the drives one by one?... i was planning on getting 4x 6tb drives instead of the 12 2tb drives i have now... but to do it replacing one by one i would need 12 6tb drives dont i before i can actually increase the size used to 6tb per drive... or is there a better way to do this without getting 12 new drives?

gptid/151b41fb-b885-11e5-b7e7-d050995afa5e N/A ada0p2
gptid/1106f45a-b885-11e5-b7e7-d050995afa5e N/A ada1p2
gptid/13258dfd-b885-11e5-b7e7-d050995afa5e N/A ada2p2
gptid/10280609-b885-11e5-b7e7-d050995afa5e N/A ada3p2
gptid/13fd7b24-b885-11e5-b7e7-d050995afa5e N/A ada4p2
gptid/121b2319-b885-11e5-b7e7-d050995afa5e N/A ada5p2
gptid/0b006576-b885-11e5-b7e7-d050995afa5e N/A ada6p2
gptid/0c9404a3-b885-11e5-b7e7-d050995afa5e N/A ada7p2
gptid/0bc9a539-b885-11e5-b7e7-d050995afa5e N/A ada8p2
gptid/e36951d1-e3d0-11e6-8703-d050995afa5e N/A ada9p2
gptid/0e8518cc-b885-11e5-b7e7-d050995afa5e N/A ada10p2
gptid/0f5a30e4-b885-11e5-b7e7-d050995afa5e N/A ada11p2
gptid/3ffe3acc-e25a-11e6-ad1a-d050995afa5e N/A da0p1
gptid/9aaf3b94-e3dd-11e6-9690-d050995afa5e N/A da1p2

The faulted drive isn't in this list but if you've already replaced it that explains it.

I'm glad you're back with an healthy pool but you should really keep an eye on the temps and please note that a 12 drives RAID-Z2 isn't recommended.

My recommendation would be to get six 6 TB drives, make a new pool with them as a RAID-Z2 vdev, copy the data from the old pool to this new one, delete the old pool and add the 12 drives to the new pool as two RAID-Z2 vdevs of 2 TB drives each. This will give you a proper setup and about 28 TiB usable (or 35 TiB without the 80 % rule) ;)

 

[Sander Jansen]

The temps is probably when i was chacking them... i have a little dock thingy and i put them in there to check them before i made the freenas setup... that dock is not cooled in any way... i dont know what else it could be... i have freenas setup that i get emails when a disk gets above 38, on average in the freenas system there around mid 30 ish...

I know its not recommended... thats why im planning on replacing them with bigger drives and keep it around max 6 drives... and probably at some point make a second pool with the other 6 slots but that is way future planning...