Hi I have been using Plex for years. I have been using my lap top as the server as well as storing my media files on it. I recently found an old hard drive that I pulled from my old laptop years ago. I decided to buy a dock for it and plug it into my router via USB. I am wondering if I could move my Plex server and media files to it so I don’t have to run it on my lap top? Or maybe there is a way that I keep the server on my laptop, but move the media files? Thanks.
A hard drive is just data storage. To function as a server, it misses the ability to process data.
You could plug in the drive to your router, which would give you a kind of improvised NAS.
But the server still has to run on a kind of computer, and your router cannot do it.
Which leaves you still the only option: run Plex server on your laptop.
Which then still requires the Laptop to be running, when you want to play something.
Where the hard disk with the media files is plugged in, in this scenario, doesn’t make much of a difference.
Okay. Thanks for your answer, so can I plug the docked hard drive into my router and just have the media server access them from there? If so, how do I go about that?
@Sunflower1980 said:
Okay. Thanks for your answer, so can I plug the docked hard drive into my router and just have the media server access them from there? If so, how do I go about that?
I cannot tell you the exact procedure. This depends a bit on your router.
If it is similar to accessing a NAS from Plex server, you set up a new library in Plex, then when it asks you about the storage location of your media files, you type in the UNC path of the router’s file share
(e.g. \\Router\share1\movies)
How the UNC path actually looks like for you depends (again) on your router. You should be able to see the UNC path if you examine the network neighbourhood with your file explorer.
I know that you already know this, Otto, but for the sake of mentioning both pro’s and con’s I’d say that plugging a USB disk into a home router has its potential drawbacks I’ts very likely that the performance of said disk will be subpar, thus not giving your Plex server data fast enough. I’ve seen it with other users who has had a USB disk (often version 2) connected to their router and they’ve reported it being sluggish and often timing out. I concurr that this subpar performance will most likely have a lot to do with other factors than a generic “router and usb” but the fact remains that most home routers have a quite “bad” implementation of this “NAS” feature.
If the aim and goal is to have a centralized storage unit (and perhaps even remove the need for your current server to stay on) I’d strongly recommend getting a real NAS or equal solution (DIY server). Sorry for this harsh “truth” but cutting corners and staying on the cheap side has it’s drawbacks and I just wanted to let you know this side of the coin.
Thanks for your answers. I know that you get what you pay for. I’ll give it a try and see how it goes. Can’t really hurt anything. 
USB is for temporary connectivity where flakiness and reliability are trumped by convenience. Anything that you want to remain connected and is important should use a proper connection.
@sremick said:
USB is for temporary connectivity where flakiness and reliability are trumped by convenience. Anything that you want to remain connected and is important should use a proper connection.
Actually I strongly disagree with this. USB is a “proper” connection. All my media drives are connected via USB and they are rock solid. They are every bit as stable and reliable as the internal drives in that server.
The reliability of USB connected drives are directly related to how the drives and ports are configured and that is no less reliable than drives that are connected in any other manner.
I have about 10 drives connect via USB (mostly USB 3.0) for every internal drive and the only drawback I see is that access via USB is a bit slower that the internal drives. The USB 3.0 connected drives are just a little slower and even the 2.0 connected ones are only noticeably slower for file transfers.
It is well to keep in mind that video serving is not very demanding as far as read/write speed is concerned. The important thing for video is that the read rate remain high enough and smooth enough so that the processor has plenty to work with and USB 2/0/3.0 rates are plenty fast enough for that.
USB connections are generally faster than network connections and we find that drives mounted on separate file servers work fine. So drives connected via USB to the server itself are faster and easier to access by the server than drives connected to another device on the network and therefore quite usable and reliable for simple file access.
Obviously it’ll vary from environment to environment, but it’s that ability to vary which leads to issues. Many USB controller chipsets are junk, and the drivers are worse. USB drop-outs, flakey drivers, and the fact that the majority of USB drives have no means to pass along SMART health data over the USB connection means you don’t discover your hard drive is going south until it’s too late and your data is lost. USB is also very susceptible to signal loss over the cable, which it doesn’t handle very elegantly.
A person could make a car (not legally) without seatbelts and with things held together with zip ties instead of bolts, and for some people during short trips for some amount of time, they might not have any issues… but it’d never be something I’d recommend to anyone no matter how many positive anecdotal experiences I hear about.
I come from the realm where a “server” is something important handling data you actually care about, which excludes the lowest tiers of technology and quality. I suppose if you don’t care about data loss and can deal with extended downtime, you can slap any type of hardware together and call it a “server” but even here among Plex users, most people care about their data and time enough to not make the little bit of savings of going to USB to be worth it. Don’t spent $100 tomorrow to save $10 today.
I will agree that USB3 can be faster than gigabit ethernet (not USB2). However, the question wasn’t so much the difference between choosing between USB and ethernet (which is meant for long distances) as USB-vs-SATA. USB introduces latency and is very CPU-heavy, as its meant for a million different purposes. SATA is meant for low-latency, high-speed transfer to/from storage devices and that’s it. And SATA3 beats the pants off USB3 in data transfer rates.
And finally: USB controllers at the drive end like to fail, either randomly or due to physical issues. Given the fact that many external hard drives (I’m looking at you, WD) are now integrating the USB controller into the drive itself, you can’t just pull the drive and mount on another SATA connector like you used to and get at your data. I don’t have enough fingers to count the number of clients who’ve lost data on their external USB drives because the USB controller died and we had no options other than sending to a data recovery place ($1000-2000). I used to just remove the drive from the enclosure and hook up to another SATA-USB adapter to read. Not anymore.
@sremick We will just have to disagree on the reliability of external USB drives. The only point of failure that I have seen that makes many USB drive less reliable than internal drives is the connector and that is primarily, I believe, because people do move the drives around a lot therefor putting extra stress on the connection.
I do not find that USB drives fail more often than internals and the USB electronics in the drives do not really have any greater chance of failure than the electronics for other connection types.
The savings of using USB drives are much greater than just the cost of the drives. Not having to invest in the expensive and fairly failure prone (no more or less than other compute equipment) hardware to enclose the drives (I currently have 12 3-5 TB drives in my setup and will need to add 2 more soon) is a savings way beyond the cost of the drives.
I use DrivePool for pooling the drives and I have duplication turned on so if a drive fails all I need do is pull that drive out of the pool and put a new one in.
Since I started with a couple of older drives I have had a couple of failures in the last years but I have lost no data from the parts of the pool where I have duplication turned on and all I had to do was remove the defective drives and plug in the new ones and inform the DrivePool software of the change.
It is possible that my situation us unique but I have had and continue to have a great experience with USB drives and I do not see a reason to change.
I do agree about WD drives with their new electronics and do not buy them since that change however I have only once needed to access data on a failed drive and I have not yet found a case where the failure was in the USB controller (except for a bad connector) when a drive went bad.
But I am sure I will never convince you, or anyone else that believes USB drive are failure prone, that they are truly just as reliable as internal drives because all the statistics include connector failures caused by moving the drives around which gives a false failure rate for drives that do not get moved a lot. If the tests were made on drive sets that were stressed the same way I am quite sure that the statistics would be different. You can prove anything you want with statistics by carefully choosing the data sets and ignoring the differences in those sets.
BTW: Cooling is quite important and I do have a fan blowing across my USB rack to keep the drives from getting too hot. This is the same as have good cooling for a case containing drives. In general USB drives are not actively cooled and in continuous operation they can get a bit too hot without good airflow around them.
