I have a Plex server setup with a dozen 8TB external drives, 2 5TB external drives, and 2 4 TB external drives. Unfortunately, my 4TB Seagate (the only Seagate I have) crashed and took my 720P content with me. That was about 1500 out of 14,000 movies, so not a terrible loss, but it had some hard to find stuff on it.
I was able to disassemble the drive and unstick the drive head. When I plugged it back in, I could access all the files. Obviously, I don’t trust the drive anymore. I am in the process of copying everything to a new 5TB external drive.
It was my “L” drive on the server, and I want to re-add it as the “L” drive, which I can handle in Windows. What I’m worried about, especially with a large library, is seamlessly reintegrating it back into the libraries. What is the best way to go about it? If I just make it the “L” drive again, will Plex pick up and keep all the metadata intact, or will I have to rescan everything? I have collections implemented and all 14,000 movies are tagged with custom tags, so I really don’t want to have do redo the 1500 or so on the drive.
If the drive paths all remain the same Plex should be able to see it and it should go back to normal. I would however strongly recommend getting away from using external drives for Plex. They can be incredibly finicky and cause some pretty silly issues. Maybe before investing in more external hard drives look at a NAS or DAS.
Thank you. I will give it a try, and hope that it works as before.
I have been running this setup for about 5 years now, and I kept meaning to move everything to a NAS. However, at this point it would be about $10,000 to get a 120TB NAS unless I can shuck the drives I have now, so I could just buy a diskless NAS. It looks like a DAS that size would be about $7500 or so. I do run drive checks on everything regularly, and replace drives if they show any sign of failing. This one caught me by surprise though, because it was just a stuck head. I still don’t trust it not to happen again though, so it’s going to store non-essential data now if I use it at all.
Using USB drives, especially in windows is no different than using a NAS or DAS. I would recommend cloning everything to a backup drive and maintain a 1:1 primary | backup configuration. Any storage medium has a lifestyle and will fail, its just a matter of when. If you are able to replace the L drive before Plex rescans and empties the trash bundles you will not see any difference as long as the file structure on the new L is the same as the old. Ive been doing this for years and have 1-2 drives a year fail. (pretty good rate given setup) I just disconnect the failed drive, re-assign the backup’s drive letter, and replace and re-populate the backup drive.
My NAS setup is 96TB and cost me about $2500 when it was all said and done. I likely could have had more storage (around 128TB) for a very small amount of extra money (probably $300-$400 more) also if I had known what I know now.
This is absolutely not true. External drives connected over USB are not going to be the same as a NAS. Two wildly different methods of access, and two incredibly different devices. The DAS argument doesn’t even really hold water as most of them use their own RAID controllers, power management, and manage the drives independent of Windows. Sure the communication ends up being the same since they are both over USB, but still not anywhere near the same thing.
To add-on to what @dbirch said about the emptying trash, I would suggest going to the main Settings for Plex (wrench icon on new versions of the Web app, wrench and screw driver icon on older, not sure when that changed). Next, under the Settings section in the left menu click on Library and temporarily uncheck “Empty trash automatically after every scan” and then go to Scheduled Tasks and temporarily uncheck make sure to uncheck “Remove old bundles every week” and “Remove old cache files every week”.
Not sure if those last 2 are absolutely required, but better safe then sorry until you get the drive replaced.
Once you replace the drive and everything is back to normal you can put those back to as they were if you want. I personally keep the one to empty trash after every scan off.
@Blkbyrd please note I was not talking about using a RAID setup, rather just a basic setup. This isnt the place to discuss RAID vs non-RAID setups. But at face value internal. USB, NAS or DAS each have their pros and cons. for a basic KISS setup USB works great there isn’t a reliance on a RAID cluster that needs fully redone, a separate NAS OS to manage, it goes back to the basic plug and play principle and can be scaled as needed without having to redo anything.
Everything is back online and running well. The metadata is there, and all the custom tags are intact. That was a close one
I have decided to do a Backblaze backup for everything. They have cheap unlimited storage. I calculated that each 8TB drive will take about 30 days to backup, and about 1.5 days to download, so once the initial backup is done for the 120TB and assuming only one drive fails at a time, it shouldn’t take long to get things back online. Thanks again for answering my questions.
