These fans have little to no power regulation, and will happily draw more power than the 5w, that a typical USB port can output. This will overheat (some) USB power circuitry, and possibly destroy it.
That atleast was the exact reason given at my university, as to why USB fans were banned
Would a small 4cm fan be ok then? Iām using it to cool a network card in a pfsense firewall where there a no internal power connectors.
How can a fan ādrawā more power than the port provides?
But this is not the reason. USB have brownout protection. The actual reason is you are connecting a very noisy inductive interruptible load to a digital power rail
What happens when current is interrupted through the coil? You get very large voltage spike. It gets clamped by protection diodes. This generates very wide spectrum noise. This happens multiple times per second in a DC motor.
And these retched plastic synology boxes lack ESD cage. Synology saved a few bucks there. 916 had it. This one does not. People report unit rebooting from ESD when you dust the shelf around it. They also I guarantee you skimped on USB power filtering and suppression ā and even if they did not, you are force feeding noise into digital power domain. Itās very dumb.
Itās a very horrible idea to connect a motor to a USB port. Donāt believe me ā measure the signal with some high bandwidth DSO.
But you may say ā but regular case fans are connected somehow? Yes. They are. To an entirely different 12V rail. To which other motors are connected. Like HDDs. Itās noisy as heck. But motors donāt care. Digital circuitry, however, does, and very much so.
So please, unplug and destroy that evil shite. I thought it was a led light. Fan has no business being on usb. It makes an EE in me physically uncomfortable.
And while we are shitting on the parade: these plastic toys are wrong synology units to waste money on. They are e-waste from the day of production. The metal units with internal PSU are actually pretty good hardware, and the the reason why nobody shall be buying them either is different from solely hardware quality.
Thatās a great explanation - thanks 
And after 5 years of running the fan in the USB portā¦ damn, is still going strong. 
So funny would be if that is also a storage node.
ā¦ the luck has almost completely ran out. Unplug.
āDepressingā is the word you are looking for, not āfunnyā.
Plastic overpriced ānasā that requires external usb Chinese cooler to function and mental gymnastics to justify.
Iām not poking at you specifically by the way. Itās a Synology thing. Bodging around cost cutting and vendor incompetence to avoid acknowledging that the overpriced turd is in fact nothing more than an overpriced turd has always been part of Synology ownership.
Good luck.
Iām under the impression that you believe that contraption is mine.
Itās just a pic sent to me by a friend, who found it on X. I also use Synologys, 2-bays only, but Iām not so stressed about the cooling part. I rely on the included fan and I trust the OS doing itās job, and just set the fan speed to Cool; not Silent, not Full. They say it will increase the speed if necesary.
Me too. No problems running usb fans ever
What are you both trying to say? Is reasoning above about why connecting motors to data ports is idiotic invalid? Point to the specific flaw in the reasoning then. Explain why.
Or you think that āI did not notice any visible issues with this my fan in this my usb port, and therefore all fans are safe on all usb ports, and therefore whatever reasoning to the contrary is provided is invalidā?
Anecdotal evidence is not justification.
Neither are two or seventy anecdotal evidence pieces. Physics is not a matter of popular vote.
Weāve gone from:
Post pictures of your storagenode rig(s)
ā¦toā¦
Post pictures of somebody elses storage
ā¦toā¦
Argue about pictures of somebody elses storage
Becauseā¦ we like to hang out and argue about storage stuff; who can name 3 friends from the phoneās contact list that could share this pleasure?
Itās okay, I can spam some of my stuff if that makes you happier? I recently turned off my old Synology SAN. Feels great being on full ZFS now!
I canāt speak for the other guy but you are just too scared of everything. According to you, usb drives donāt work, PC motherboards donāt work, storj doesnāt work, docker doesnāt work, usb fans donāt workā¦no doubt āwall wartsā donāt work either.
You can assemble a vast argument about why things donāt work and then completely ignore the actual evidence that all of these things actually do work.
Science does not exist just in your mind; It was to also make useful predictions and when those predictions fail, like all your predictions have, you have to reassess your ideas
I really enjoy seeing what other people are running. Itās often things that Iāll never manage to own myself: some homelabs are amazing!
Iām not scared of bad design and mediocre engineering, I despise it, reject it, and donāt make excuses to keep tolerating it.
I reject the disgusting āgood enoughā approach outright and explain why, even though I donāt have to.
āworks for me, ship itā position allows mediocrity to continue exist.
You have listed a lot of technologies in a bucket list that are out of scope of this thread.
Focus on usb fan. If you have technical objection - go for it. Otherwise quit normalizing mediocrity.
If you want to discuss docker and wall warts ā ask specific questions.
You are calling anecdotes āactual evidenceā and then inflating them into a general rule. You are living the smoking at a gas station example again.
āMy USB fan workedā is evidence that your USB fan worked in your setup. It is not evidence that using USB data ports as motor power supplies is good engineering.
The argument was not āeverything fails immediately.ā The argument was that this is the wrong way to peer that type of load.
That point still stand, despite of what you observe.
You are treating ācatastrophic failure did not happen to meā as experimental proof. That is not science. That is a small sample set with a overconfident conclusion.
What prediction are you referring to?
I did not predict that every USB fan immediately kills every USB port or always injects data bugs straight into the ram. I explained why powering a motor from a data port is bad design.
The correct response to bad design is to stop doing it, not to defend the abomination because it happened to work.
Go read gas station reference above again, it answers all your questions.
Just turned off?
Ok sorry about the attack.
Seriously though, I said I havenāt had any issues with usb fans. What I really meant was Iāve seen tens of usb fans in use. I have friends who use tech. I go to offices. I see dead people ā¦ oh not that one! I have never ever seen anyone having any problem with a usb fan.
I donāt have much reason to keep coming here to be honest but I have read your posts and rate them so as the people say - you do you -
Iām still waiting for the last zfs attach to finish, then the final files will be flushed from the remaining disks in the rs3617xs synology.
It will stay for a while; Iāve grown quite attached to it, and the many modifications Iāve made to it.
Will probably sell on used marked to recoup cost ASAP, and get on with life ā¦ or try to mess with itās sataDOM and boot it to an alternative operating system
Selling nowā¦ may be the best price you can get. And no matter how attached I was to gearā¦ after I sold it I never regretted it. Thereās always some new shiny thing to spend money on 
