Plex RAM Transcoding, Better than an SSD?

preview_player
Показать описание
After many told me to do it, I finally stop ignoring and make the change. But did it make it faster?

*Advertisement Info*
Get up to 96% OFF on 1, 3 & 5 year .TECH domains -
Use code - BMB - at checkout
Make sure to select 1, 3 or 5 from the dropdown menu

Want a secure and fast VPN? Hate Netflix Geolocks?
Use my link to help me out :)
----Use the coupon code "byte" to get 77% off of three years!

*Plex Affiliate Links*

Support my growth!

Twitter:

*📷This is the gear I use🎥*
📷This is the gear I use🎥
Рекомендации по теме
Комментарии
Автор

IMO Plex Server should have a settings option to utilize all available RAM for transcoding/PVR playback.

ITSALLFYA
Автор

I just changed transcoding to /temp folder and I can say that it makes difference. I have 16GB DDR3 1600 memory. Transcoding starts faster and seeking time is also much faster. I can actually play content through Firefox. That is something I was not being able to do before due to error you described before. Thanks for this tip

talktomenowxbmc
Автор

In Linux just use /dev/shm as your Transcode path under Server Settings. Done.

jonathan.sullivan
Автор

Jason you won't likely hit this limitation with 128GB, but to those who try this remember that the Plex Transcoder does a free space check before it begins to transcode. I believe it checks that free space is equal to the size of the file being transcoded + 10%. If you get random crashes after doing this, this is probably why.

camaromike
Автор

I found a noticeable difference running the Plex folder on an SSD and the transcode folder on a RAM drive. The biggest plus is video tracking, it recovers much quicker to continue playback.

frankleelost
Автор

Good info. Thanks man. I added a RAM disk and passed it through to my PMS from Proxmox in just a few minutes. I didn't realize how many files were written during a transcode. Speed be damned, this has to help my SSD's longevity.

maltoNitho
Автор

I've been using tmpfs for my Plex transcode target for a year or 2 now, and it works flawlessly, even on my mediocrely-powered gen8 HP Microserver.

There's a caveat to this, though. You have 128GB RAM, so not a problem for you, but for everyone else who might be running on a more limited amount, if your tmpfs fills up during a transcode session, ALL streams will stop, with an error written to the log file saying it can't write the temporary files due to lack of disk space. Plex does NOT clean up until AFTER the transcode session has ended (to allow for faster seeks to areas it's already transcoded). This means that, if your tmpfs isn't large enough and you're transcoding to a format that will ultimately end up with a large amount of data generated (or you have multiple streams on the go), your (and all) stream(s) will stop when your /tmp fills up. It's at this point that your /tmp will be emptied, and you'll be able to restart the play/transcode session from where you left off. It's only a minor thing because it remembers where it stopped, but annoying nonetheless.

Mine is set to 8GB (because I only have 16GB RAM in my machine) and this is adequate for most of my streams, but then a lot of what I watch is done via direct play, which doesn't seem to affect it as much. In other words, the more RAM you can spare for your tmpfs, the less likely you are to fill it up. But if you only have like 4/6GB RAM in your machine, you likely won't be able to allocate enough memory for a tmpfs location AND keep your machine from choking due to lack of available RAM.

dannosaur
Автор

I found this interesting. I use Plex as a VM on HyperV, server 2012. Standard spinning disks in Raid. Transcoding a bluray or switching transcoding modes (720, 480 etc) is almost instant maybe 2 seconds max. 5 seconds sounds kind of long when you're throwing as much power at is as you are. Cool experiment though, but seems unnecessary.

technologyunloaded
Автор

SSD is a great storage media. The only issue is that it has limitations on how many times you can rewrite or re-strip the draw.

Meanwhile, Ram is literally built to be cache for your system.

jarek
Автор

Samsung 850 Pro line, built with 3D V-NAND. Capacities include 128GB, 256GB, 512GB; and 1, 2, and 4TB. Samsung calculates the TBW by drive capacity. It offers a 10-year warranty if the writes are less than TBW maximums.

256 GB: 150 TBW
512 GB and 1 TB: 300 TBW
2 TB: 450 TBW
4 TB: 600 TBW

So basically you got a dud SSD that failed no way you were even close to the expected life.

NextLevelCode
Автор

A typical TBW figure for a 250 GB SSD lies between 60 and 150 terabytes written. That means, to get over a guaranteed TBW of 70, a user would have to write 190(!) GB daily over a period of one year. (In other words, to fill two-thirds of the SSD with new data every day). In a consumer environment, this is highly unlikely.

blazerdel
Автор

The other option is to use an Intel Optane drive for your transcode drive. (not the 16 or 32gb versions, but a 900p or 905p) I have a 900p 280gb version used for transcoding. Good thing is they are extremely fast and low latency due to being NVME and also they can handle extreme number of writes. Another option is an Intel P3700, which has a 5 year life span with over 17 drive writes per day.

CharlieRasch
Автор

do you still happen to have the failed sandisk drive? i'd be curious if the sandisk software on windows could see just how much data had been written to the drive prior to failure...either way hope Zeus stays strong for you! P.S. do you have turbo boost enabled through the tips and tweaks plugin? that could give you a little bit of an uplift

codeman
Автор

I never thought of this. You, sir, win today.

StevenLynn
Автор

I guess it doesnt make a difference in speed, but it make a difference in writes/reads on hdd. And i guess, more then one transcode could be faster on RAM.

galsherp
Автор

Can you please do a freenas version of this ?

chadhelou
Автор

Hey Jason! I know this is 3yrs ago but curious if this is holding up for the long term? Have you had any issues with ram failing or anything like that? Do you have any idea how much ram you typically use when watching or transcoding 4k bluerays? Does this help take off cpu load when watching 4k bluerays? I can play 4k fine but it runs my cpu around 50%....I'm wondering if running off of ram will reduce that some. Either way, I'd like to try this out.

stephenwhite
Автор

I thought I remembered them removing this as an option at one point in time. I have now swapped back! Thanks for the video :)

popple
Автор

Hi Noob question when having this arrangement does the Ram self clear after transcodes completed/movie has stopped playing? so the Ram doesn't fill up

lukehodgkinson
Автор

I would have to disagree with the Sandisk comment. I have 12 1TB Sandisk Ultra II's running in a Ceph cluster, from which I run about 35 LXC and 3 QEMU instances. The drives in the cluster have been going strong for nearly 2 years now, and I also have another 1TB Sandisk Ultra II in a USB 3.1 external enclosure for transportable storage. I've had a great experience with Sandisk and recommend them when asked. They'll all eventually fail, but I expect it to be outside of the warranty period.

tycoonbob