firedfly

That's the plan. The best way to defeat @BWG is to use his own farm against him! Thanks!

In firedfly That folding time prize of 1xBWG's farm is quite the prize!

I just looked at your screenshots. My first thought was that you have a LOT of drives only covered from the 1 parity drive. If you lose 2 of those older, smaller drives, you'll start losing data. Though, I'm definitely on the more cautious side (2 parity/9 data drives). As for the cache drive not being used, you would need to check your share settings. I have Use cache pool set to yes for the shares I want to utilize the cache (and have the files moved to the array when the mover is invoked). Also, if you are interested in my configuration, I have multiple cache pools. One for normal array caching. The other stores my docker and vm files so they are not on the array. Additionally, I'll pass some specific SSDs through to a few VMs when a dedicated SSD is preferred.

I use unRAID and like it a lot. I don't know anything about truenas. I feel the biggest drawback to unRAID is access speed when accessing the array... Specifically writing to the array. Read speeds are capped at the speed of the drive hosting the data you want. Unlike other systems, you won't get the speed benefit of reading or writing to multiple drives at the same time. As for writing to the array, it is slow by default. Using cache drives to receive new or updated files helps, but can have issues (writing more data in a day than the cache drive can hold). There are other settings to speed up writes if you keep the drives spun up at all times. Overall, I love unRAID. However, I have a slew of non array ssds to keep things fast.

I have my Plex server running on my Unraid system. The ramdisk is 8g. I generally cap out at about 3 simultaneous streams, so I'm not sure how much space would be needed with more streams. As I mentioned before, my only issue with using the ramdisk was when users would pause what they were watching. Transcoding would continue and fill the ramdisk. I changed some settings a long time ago and have not seen that issue in quite a while. The only setting I remember changing is killing a session after it has been paused for a set amount of time. This is in the network settings: For anyone else running Plex on their Unraid system, I have a user script setup to create the ramdisk when the array first starts. This is the script. You'll need to make sure to update the Plex docker settings to use the ramdisk for the transcoding directory.

Not this time...this is a Linux system. Must have been a power outage and fah didn't start back up properly when the power came back on.

I've done this. I had an issue in the past where the ram disk would fill up. The issue was user's pausing and, I guess, transcoding continued. I fixed this by increasing the size of the ram disk and configuring Plex to kill sessions that have been paused for X minutes (15 min?). I also have a gtx1660 handling transcoding duties. It works quite well and greatly minimizes the load on the CPU.

Just discovered my ETF rig has been down since May 6. It should be running again now.

I notice this frequently on my Android phone using Chrome. The page auto-scrolls past the oldest unread comment and I have to scroll back up. Using Chrome in Windows, I've not noticed this issue.

Thanks! I did pick up an upHere model support bracket, though a different one from what you linked to. It looks like it will work well.

I didn't expect you to find out this quickly!

I didn't think it was happening until later tonight. I just refreshed my EOC stats page and @Bastiaan_NL had disappeared!

I'm a fan of a mesh back. However, I won't get another chair with a mesh bottom. It was fine at first, but over the course of a few months, the mesh seemed to stretch a bit and I was eventually sitting on the frame around the edges of the seat bottom. It became quite uncomfortable. Having said that, that type of chair (mesh bottom/back) is the default that my company purchases for employees. I don't recall anyone else complaining about that issue. Perhaps I got a dud...or just need to lose some weight

IN firedfly

I asked my wife to give me a name for something that is ancient, slow and yet powerful. She pretty quickly came up with Master Oogway from Kung Fu Panda. It seemed very fitting.

Got an amazon link to your preferred bracket? I'll buy it now if so. Otherwise, I'll never get around to picking one EDIT: Also, the sag look on the picture where I added the 3070ti isn't sag. The system was laying down at that time. I still think an anti-sag bracket is a good idea, though!

@ENTERPRISE It looks like this was an issue on my side when adding the images. You can ignore this issue. @Avacado Are the images working for you now?

I'll try to redo the images. I just checked one of the ones I can see and it shows as pulling from google. Perhaps I did mess up when adding them. I had copied the pictures from Google Photos and pasted into ehw. I *thought* that copied the actual image, but perhaps it just copied a link?

@ENTERPRISE Do you have any guidance on how to resolve the broken pictures mentioned above. I see the pictures fine and even tried on a second computer in a different location and the pictures displayed properly.

Agreed. I find this build quite funny. Itx board in a full case. Giant video card installed on the tiny board.

For the last few weeks, I've had a 12gb 3080 and a 3070ti (both EVGA FTW3) sitting in boxes waiting to be put to use. My original plan was to upgrade my main desktop from the gtx960 that it has been running for years. I'd never make use of either card while gaming, but my desktop was a good option for housing both GPUs for folding...or so I thought. I finally installed the 3080 a few days ago and realized that while the 3070ti would also fit in the case, the the top card would not have any breathing room to keep cool. Unfortunately, this left the 3070ti without a home. Digging through my limited supply of old hardware, I came across an old HTPC from around 2011: ASRock H67M-ITX motherboard (yes, a tiny itx board) i3-2100t (2 cores) 2x2gb DDR3-1333 ram This motherboard has a single PCI-e 2.0 x16 slot. Finding out how old (and tiny) this system was, I started having doubts that the 3070ti would work, much less fold. I cleaned out the decade of dust: Installed the tiny ITX board into a spare case: And, finally, installed the 3070ti: There were some issues booting with the 3070ti outputting to the display, but the onboard video worked around the issue. Soon, I was presented with an old copy of Windows 7! It had 4gb of space free on the 60gb ssd. I uninstalled a slew of old software that wouldn't be needed and to free up some extra space. Nvidia drivers for Windows 7 installed without a hitch. I didn't even know they still released drivers for Windows 7. Folding@Home also installed without issue. It wasn't too long before this ancient system with a new GPU was folding. I'm quite impressed that this system is able to fold on the 3070ti. I'm even more impressed at the performance. I have another system with a 3070ti (currently not powered on) that will get about 5.4 million ppd in Windows 10 on p18201. This old windows 7 system with a dual core processor and pci-e 2.0 is getting about 4.4 million ppd! I'll eventually re-work some of my other systems to move this 3070ti to a faster, more capable system for better folding performance. But for now, Oogway will fold on!

@Cerberus It looks like your ETF card is down.

Do you have an ETA on getting your card back online? I just realized it was still down. If it is going to be a while longer, I should be able to cover for a bit with an AMD card.

@Bastiaan_NL I need to add a second 3070ti (08G-P5-3797-KL) to my list. I stepped up to it on 2/16/22 from a 3050 (08G-P5-3553-KR).

Not the same card, but I have two air cooled 3080TI FTW3 cards (12G-P5-3967-KR). While folding, I'm seeing temps of 60-65C with 75F ambient. I also have a hydro copper (12G-P5-3969-KR) that is currently 47C in that same system (75F ambient) while folding. All three of these are in an open frame case with a lot of airflow.