So realistically, how bad is 48-50C only during parity checks

[tyrindor]

I have a Norco 4224 case with the optional 120mm fan bay. I have all Noctua fans running at full speeds, however, the hotswap bays don't allow much airlflow. I can take one out and blow into it, and I barely feel the air on the other side. I can set up a box fan on max speed 2 inches in front of the hotswap bays and the drives don't even drop 1C.

 

This was never really a problem until I upgraded to 8TB Seagate Archive drives (they run about 5C hotter than the WD reds they replaced), and I have 24 of them in there. The drives stay below 40C outside of parity checks, but for ~18 hours a month they will be at 48-50C. The drives are rated for up to 60C. I don't think even putting 3000+RPM "jet engine" fans in there would improve it with the current hotswap bays.

 

Should I be concerned? Surely if the drives are rated to 60C, running them only 18 hours a month at 50C won't hurt lifetime a noticeable amount?

[Squid]

Going to leave comments on the actual temps to others more knowledgeable.

 

But, the problem with cooling in hotswap cases is that you need high static pressure fans (like the ones that came with the case).  And high static pressure fans generally means loud (which the noctua's are not)  And high CFM does not equal high static pressure (which is required to pull air through a very small crack)

[JonathanM]

I wouldn't worry about it. They are rated for continuous use at max 60C, so brief periods at 50 are well within design specs.

[SSD]

I have a Norco 4224 case with the optional 120mm fan bay. I have all Noctua fans running at full speeds, however, the hotswap bays don't allow much airlflow. I can take one out and blow into it, and I barely feel the air on the other side. I can set up a box fan on max speed 2 inches in front of the hotswap bays and the drives don't even drop 1C.

 

This was never really a problem until I upgraded to 8TB Seagate Archive drives (they run about 5C hotter than the WD reds they replaced), and I have 24 of them in there. The drives stay below 40C outside of parity checks, but for ~18 hours a month they will be at 48-50C. The drives are rated for up to 60C. I don't think even putting 3000+RPM "jet engine" fans in there would improve it with the current hotswap bays.

 

Should I be concerned? Surely if the drives are rated to 60C, running them only 18 hours a month at 50C won't hurt lifetime a noticeable amount?

 

These kinds of questions are pretty impossible to answer. It comes down to a question of statistics - and without a fairly large statistical sampling resembling your usage pattern, it is in the world of conjecture. I have made a study of drives and heat, and best information is that operating a drive at a higher temperature is not as bad as operating a drive with a large temperature differential between its low temp and high temp. So if your drives are normally at, say 40C, and you run a parity check and they hit 50C - that would be better than having your drives idling at 30C, and having them hit 50C on a parity check.

 

I have personally tried to always keep my drives under 45C, and have had very good luck with long lasting drives. I know from experience that heat kills electronics, and figure the cooler the better (although if you read you'll learn that drives kept too cool negative affect their longevity).  If it were me, I'd pop in beefier fans. But my solution to maintaining a completely silent array is locating it in my basement where I can't hear even the noisiest fans. But I'd say your add'l risk is pretty low hitting 50C for parity checks. But you have a huge investment in drives (24*~250 = $6000), and adding a bit more cooling might mean fewer replacements over the next 4-5 years. Either way, you'll never kmow if you made the right choice.

 

 

[garycase]

While 50 is not a dangerously hot temp, it is a bit higher than I'd like to see.  But probably not an issue since it's only hitting that range for 18 hours/month.

 

It'd be very interesting to see how much that could be improved with Noctua's industrial quality NF F12 fan [ http://noctua.at/en/nf-f12-industrialppc-3000-pwm/specification ], which is designed for high static pressure (7.63 mm H2O) -- far more than the consumer grade NF-P12 (1.68 mm H2O)  or  NF-F12 (2.61 mm H2O) units.    I suspect that would knock several degrees off the temps, but whether or not it's worth it to do that is debatable when it's only an issue for such a small percentage of the operating time.

[ashman70]

I would suspect it might affect the overall health/life of the drive, but I don't have any data to prove it, just my gut a experience.

[TUMS]

I've got a norco 4224 and did a parity check yesterday and my disk temps never got over 35C. Of course its winter, my summer temps aren't near as good.  I'll be switching to a supermicro chassis soon and I'm curious to know how my temps will be just switching cases.

A lot of times in the summer I will turn a big fan on and point it towards the front of my server case. It helps quite a bit. You might want to try that OP, at least during parity checks.

[Helmonder]

I am running with 13 drives in Supermicro 5in3 cases with silent noctua fans.

 

I have been running a parity sync AND a data drive rebuild at the same time for 12 hours now. Temps are below 35 degrees with a few drives at 37 degrees.. These are allmost all WD RED's.

 

The two parity drives are 39 and 40 degrees, but these are Seagate desktop drives with higher speed.

 

My ambient temperature is 21 degrees.

 

So it can be lower.. But I would not worry..

[tyrindor]

While 50 is not a dangerously hot temp, it is a bit higher than I'd like to see.  But probably not an issue since it's only hitting that range for 18 hours/month.

 

It'd be very interesting to see how much that could be improved with Noctua's industrial quality NF F12 fan [ http://noctua.at/en/nf-f12-industrialppc-3000-pwm/specification ], which is designed for high static pressure (7.63 mm H2O) -- far more than the consumer grade NF-P12 (1.68 mm H2O)  or  NF-F12 (2.61 mm H2O) units.    I suspect that would knock several degrees off the temps, but whether or not it's worth it to do that is debatable when it's only an issue for such a small percentage of the operating time.

 

Those are the fans I was looking at getting, currently have NF-P12s in there. The problem is running 3 of those at max speed would be like a jet engine. I'd need to install autofan control or something.

 

I am running with 13 drives in Supermicro 5in3 cases with silent noctua fans.

 

I have been running a parity sync AND a data drive rebuild at the same time for 12 hours now. Temps are below 35 degrees with a few drives at 37 degrees.. These are allmost all WD RED's.

 

The two parity drives are 39 and 40 degrees, but these are Seagate desktop drives with higher speed.

 

My ambient temperature is 21 degrees.

 

So it can be lower.. But I would not worry..

I've got a norco 4224 and did a parity check yesterday and my disk temps never got over 35C. Of course its winter, my summer temps aren't near as good.  I'll be switching to a supermicro chassis soon and I'm curious to know how my temps will be just switching cases.

A lot of times in the summer I will turn a big fan on and point it towards the front of my server case. It helps quite a bit. You might want to try that OP, at least during parity checks.

 

I am running 24 drives in there, with 2 SSDs for cache inside. The archive drives run pretty hot in general. 4x WD Red the max I saw was 42C.

 

I've owned 3 different 4224 over the years, and each time the hotswap bays were different. The current V2 case I have (I believe one version less than the newest SATA6 version) has pretty terrible hotswap bays compared to the V1 case I had awhile ago. The old version had much bigger grills in its hotswap bays, and even had a switch to turn on/off airflow to each individual bay. No idea why they got rid of that.

[garycase]

...

Those are the fans I was looking at getting, currently have NF-P12s in there. The problem is running 3 of those at max speed would be like a jet engine. I'd need to install autofan control or something.

 

They would indeed be notably louder ---  43db vs 20db for your p12's.  However, they're also PWM fans, so they could be automatically controlled with a temperature-based fan controller ... or you could simply crank them up when you do a parity check.  [New feature request:  automatic fan speed ramp-up on parity checks !!  ]

 

They DO, however provide about 4.5 times the static pressure of the PF-12's ... which HAS to make a significant difference in the effectiveness you'd see.    I wouldn't be at all surprised to see at least a 5 degree drop in temps.

 

[garycase]

So the real question is do you want to spend $75 to give these a try?

 

https://www.amazon.com/Focused-NF-F12-iPPC-3000-PWM/dp/B00KFCRATC/ref=sr_1_1?ie=UTF8&qid=1483892316&sr=8-1&keywords=noctua+3000

 

Personally, I think they're almost certainly worth it -- with that much increase in static pressure there simply has to be a big gain in cooling efficiency.  The noise issue is probably pretty easy to control ... in normal (non parity check) operation you could set them to run at around the same rpm's as your PF-12s;  and during parity checks ramp them up to a high enough speed to keep the temps under 45  (You likely don't even need full speed for that).

 

[SSD]

A little off topic - but while talking about techniques to extend drive life - I have my spin down timer set at 5 hours. I think the default is one hour, and when set to that time, I see my drives spinning up and down multiple times a day - with operating temps similarly going up and down multiple times a day (especially weekends when I am actively using my systems on and off throughout the day). 5 hour spindowns tend to result in a drive only spinning up once and staying spinning all day.

 

YMMV.

[c3]

I have a Norco 4224 case with the optional 120mm fan bay. I have all Noctua fans running at full speeds, however, the hotswap bays don't allow much airlflow. I can take one out and blow into it, and I barely feel the air on the other side. I can set up a box fan on max speed 2 inches in front of the hotswap bays and the drives don't even drop 1C.

 

This was never really a problem until I upgraded to 8TB Seagate Archive drives (they run about 5C hotter than the WD reds they replaced), and I have 24 of them in there. The drives stay below 40C outside of parity checks, but for ~18 hours a month they will be at 48-50C. The drives are rated for up to 60C. I don't think even putting 3000+RPM "jet engine" fans in there would improve it with the current hotswap bays.

 

Should I be concerned? Surely if the drives are rated to 60C, running them only 18 hours a month at 50C won't hurt lifetime a noticeable amount?

 

All 4224 are not equal. Depending on the version you have, and you changed the mid-plane, there are some mods for proper airflow. The air is drawn across the drives. This means the fans need to be a suction, not static pressure, design. Hence your external fan does not help. You need to take care to ensure the air reflow around the mid-plane is minimized. The mid-plane needs to be sealed to make sure no air from the pressurized side is short cutting into the space between the drives and fans. Back in the day, there were socks inserted into the cable holes of the mid-plane. Lastly you can go the extreme and put a foam strip along the top of the mid-plane so the cover (which must be in place for proper air flow) tightly seals to the mid-plane. Any screw hole forward of the mid-plane on the bottom or sides should be taped over. Any empty drive bays need to have fillers, again socks have been used.

 

I use the rather noisy, but powerful delta fans on the mid-plane, and make sure the CPU section has plenty of exhaust venting. I do not use exhaust fans on the rear panel.

 

The 8TB are hotter because they are larger, which reduces airflow.

[tyrindor]

^ Thanks I will try sealing up some holes.

 

...

Those are the fans I was looking at getting, currently have NF-P12s in there. The problem is running 3 of those at max speed would be like a jet engine. I'd need to install autofan control or something.

 

They would indeed be notably louder ---  43db vs 20db for your p12's.  However, they're also PWM fans, so they could be automatically controlled with a temperature-based fan controller ... or you could simply crank them up when you do a parity check.  [New feature request:  automatic fan speed ramp-up on parity checks !!  ]

 

They DO, however provide about 4.5 times the static pressure of the PF-12's ... which HAS to make a significant difference in the effectiveness you'd see.    I wouldn't be at all surprised to see at least a 5 degree drop in temps.

 

Can I use the dynamix autofan plugin to turn them up/down based on HD temps? or would I need a fan controller too?

 

So the real question is do you want to spend $75 to give these a try?

 

https://www.amazon.com/Focused-NF-F12-iPPC-3000-PWM/dp/B00KFCRATC/ref=sr_1_1?ie=UTF8&qid=1483892316&sr=8-1&keywords=noctua+3000

 

Personally, I think they're almost certainly worth it -- with that much increase in static pressure there simply has to be a big gain in cooling efficiency.  The noise issue is probably pretty easy to control ... in normal (non parity check) operation you could set them to run at around the same rpm's as your PF-12s;  and during parity checks ramp them up to a high enough speed to keep the temps under 45  (You likely don't even need full speed for that).

 

Yes I would be willing to spend the $75 as long as I know they'd work with the dynamix autofan plugin. I asked a question in that thread but so far no response. I'm not sure how that plugin works, can it control PWM fans connected directly to the mobo or do I need a fan controller too?

 

Basically, just gotta figure out how I will turn them up/down as I do not have a fan controller.

[JorgeB]

Yes I would be willing to spend the $75 as long as I know they'd work with the dynamix autofan plugin. I asked a question in that thread but so far no response. I'm not sure how that plugin works, can it PWM fans connected directly to the mobo or do I need a fan controller too?

 

Basically, just gotta figure out how I will turn them up/down as I do not have a fan controller.

 

If works with some boards, including yours if your sig is up to date.

[garycase]

Excellent idea -- I don't use that plugin ('nor do I have the same board), so I can't confirm whether or not it works => but fortunately Johnnie was able to confirm that.

 

So you've already got a nice plugin that will ramp the fans down to "silent" most of the time; and ramp them up when the temps start to rise.

 

One thought:  I'm confident Johnnie is correct, but there's a trivial way to absolutely confirm it ... just set it up with your current fans 

[JorgeB]

Before the current one I used that board in my main server and it worked, IIRC the fan control is dived in two groups, fan 1-4 and fan A, you can only control one of the groups, I used fan A connected to two fans with a splitter as it was a relative small server with only 8 disks, remaining fans were controlled by the bios.

[tyrindor]

Before the current one I used that board in my main server and it worked, IIRC the fan control is dived in two groups, fan 1-4 and fan A, you can only control one of the groups, I used fan A connected to two fans with a splitter as it was a relative small server with only 8 disks, remaining fans were controlled by the bios.

 

Installed the plugin, I see nothing under "PWM controller" but I also have no PWM fans connected. Is this normal? Would like to make sure before I drop $80 on fans.

 

Sadly don't have any PWM fans in the house.

[JorgeB]

It's been some time but you'll need to run the sensor detection first, there's a post with the instructions on the  dynamix plug-ins thread.

[tyrindor]

It's been some time but you'll need to run the sensor detection first, there's a post with the instructions on the  dynamix plug-ins thread.

 

I've looked for a good 20 minutes now, perhaps I am blind.

 

All i've found is I need to install perl (no clue how) and not much documentation anywhere... seems like this should be neatly written on the first post.

 

EDIT: Installed the nerdpack to get perl, then installed dynamix system temp to get the latest sensor package.... I show 3 things in autofan "nct6776 - pwm 1", "nct6776 - pwm 2", and "nct6776 - pwm 3". I'd assume I need fan adapters to connect it to a single PWM connector? This right?

[garycase]

.... Sadly don't have any PWM fans in the house.

 

    I thought your NF=P12's were PWM, but just noted there are both PFM and non-PFM versions of them.

 

 

 

[tyrindor]

Found a PWM fan, got it working. Gonna order them and see how it goes.

[JorgeB]

EDIT: Installed the nerdpack to get perl, then installed dynamix system temp to get the latest sensor package.... I show 3 things in autofan "nct6776 - pwm 1", "nct6776 - pwm 2", and "nct6776 - pwm 3". I'd assume I need fan adapters to connect it to a single PWM connector? This right?

 

IIRC only 2 of those work, one will simultaneously control fans 1-4 and the other fan A.

[tyrindor]

EDIT: Installed the nerdpack to get perl, then installed dynamix system temp to get the latest sensor package.... I show 3 things in autofan "nct6776 - pwm 1", "nct6776 - pwm 2", and "nct6776 - pwm 3". I'd assume I need fan adapters to connect it to a single PWM connector? This right?

 

IIRC only 2 of those work, one will simultaneously control fans 1-4 and the other fan A.

 

Ah so I don't need this 3 to 1 PWM fan adapter I bought?

[JorgeB]

Only if you want to connect more than one fan to fan A since it's alone in a group, the other 4 are on the other group.