Smart tempature errors on the same disk

[Clayton]

howdy!

Quick question. I have an smart error that comes up fairly often and I'm thinking its a problem with the drive. My question is has anyone else experienced this and is there a way I can prompt WD to replace the drive under warranty hopefully before failure.

I set my smart drive tempature warning to 40c as informational and 45c as critical. Of the 8 drives in my SilverStone DS380B case, the only drive (ADA1) seems to have a high temperature. This same drive also vibrates more then the others as I can feel it though the hotswap catty. The position of the drive is the second from the bottom of the case. I bring the forums attention just in case its an air flow issue with the case design as I have heard some people say the airflow is not the greatest. That is why I wanted to know if anyone else has seen this.

 

[Bidule0hm]

"is there a way I can prompt WD to replace the drive under warranty hopefully before failure." Seriously?

It's up to you to achieve the proper cooling of your drives, it's not the constructor fault and it'll never replace a drive under warranty because of a high temp on it.

So, stop your server, rethink the cooling system (fan profiles in the BIOS or IPMI, number and/or placement of fans, type of fans, ...), do the proper changes and then restart your server ;)

 

[Clayton]

Thanks for the constructive information.

while it is possible the excess heat is due to improper cooling... I beleive it is the specific drive itself generating more heat then the others. my reasoning being that the drive in question also vibrates more then the other 7. they are all 4GB red drives so I would assume they should run the same temperature unless the case is not cooling properly/evenly. its just odd that it would be on the bottom and not on the end and experience a heat difference of 5c or more to every other drive. heat typically rises so I would thing the drives above would be hotter but none ever are.

heat alone I know WD would probably not do anything but if its something that constant are they any metrics that I would be able to show to WD that might warrant a replacement before the drive fails from other heat related factors.

 

[Bidule0hm]

4GB? you mean 4TB?

Vibrates more? how you can know that? The drives should be firmly fixed on something solid and heavy (usually a metal cage itself fixed on the metal case) so they shouldn't vibrate ;)

You can check the SMART attributes of the drive to see if something is wrong ;)

 

[Clayton]

I have a SilverStone DS380B case which which has 8 hot swap bays. Each bay has a plastic catty which holds the drive. The catty slides into a slot so there is a little wiggle room. If i place my hand on the catty face i can feel the vibration.
Ill check smart and see if anything stands out.

 

[Bidule0hm]

Yeah, that's why I don't like at all the hot swap bays.

So, the thing is, without putting an accelerometer on the drive you can't really tell if a drive vibrate more than another because of the placebo effect (unless it's very very obvious) :)

 

[marbus90]

45C isn't critical in general. However my 7200rpm drives are below that - and they don't get any kind of airflow...

 

[Black Ninja]

howdy!

Quick question. I have an smart error that comes up fairly often and I'm thinking its a problem with the drive. My question is has anyone else experienced this and is there a way I can prompt WD to replace the drive under warranty hopefully before failure.

I set my smart drive tempature warning to 40c as informational and 45c as critical. Of the 8 drives in my SilverStone DS380B case, the only drive (ADA1) seems to have a high temperature. This same drive also vibrates more then the others as I can feel it though the hotswap catty. The position of the drive is the second from the bottom of the case. I bring the forums attention just in case its an air flow issue with the case design as I have heard some people say the airflow is not the greatest. That is why I wanted to know if anyone else has seen this.

The same exact think happened to me. What I did I set the temp from 45 to be notified if it reaches 50 and the problem is solved. I wouldn't worry about hdd temps under 50C ( according to Google white paper based on Gazillion hdd drives ).
For vibration part - I would worry about it if pool has no redundancy

 

[Bidule0hm]

You should look more closely to the google paper (especially the 'Figure 5: AFR for average drive temperature"), if the drive temp is over 40 C there is a very significant increase in the failure rate (especially for drives that are over 2 years old), let alone > 45 C :)

 

[Black Ninja]

You should look more closely to the google paper (especially the 'Figure 5: AFR for average drive temperature"), if the drive temp is over 40 C there is a very significant increase in the failure rate (especially for drives that are over 2 years old), let alone > 45 C :)

Not if passe 3 years then is still good. ))))))))))))))))
Here is the Figure for everyone to interpreter himself.;)

 

[Bidule0hm]

Even after: at 4 years you jump from about 2.5% for 30-40 C to about 5% for > 45 C, so it's a factor of 2 :)

But everyone do whatever he wants to do of course, it was just an advise ;)

 

[Black Ninja]

Even after: at 4 years you jump from about 2.5% for 30-40 C to about 5% for > 45 C, so it's a factor of 2 :)

But everyone do whatever he wants to do of course, it was just an advise ;)

Well his system is NEW , for new hdd what temp you would suggest ?:D

 

[Bidule0hm]

35 C idle, 35 C loaded :P (mechanical systems love constant temp :))

Well, more realistically: 35 idle, 40 loaded ;)

 

[Black Ninja]

35 C idle, 35 C loaded :p (mechanical systems love constant temp :))

Well, more realistically: 35 idle, 40 loaded ;)

Well the study show from 35-40 , so really we can't say 40 to be better than 35 because they are in the same group. I think we are pretty much on the same page with you. Initially what I was trying to say is that 45C is not a temp that should be dealt with , and to set it under alarm over 50 so it wont be bothered with notification every time it reaches 45C because it's not an issue that need fixing. 50C it's on the warmer end , but still reasonable temp. There is no perfect temp. My SSD goes to 45C+ under load.

 

[Bidule0hm]

"Well the study show from 35-40 , so really we can't say 40 to be better than 35 because they are in the same group." yeah but if you trace a curve passing by the AFR % for each year group of bars on the histogram you can see a sort of parabole so you can extrapolate the best temp from the lowest point of the curve for each year (or do the average of the curves before and then extract the best average temp) :D Not utterly scientific but it's the best interpolation you can use here and it's not that bad I think ;)

Well, SSD are a whole another thing. Basically the temp is far much less important (not important at all?) if it stay between about 0 to 50 C :)

 

[Black Ninja]

I couldn't find anything I could disagree with you , so I do agree on 100%:)