rm414

🥴 The old enclosure fans used molex connectors and the new one used PWM. I had left the molex unplugged from the PSU, forgetting that it also powered the disks with the breakouts. Thanks for the pointer that led me to my boneheaded mistake!

unraid-diagnostics-20230630-1011.zip

I've been running my system for years without issue, but I just moved everything over to a new enclosure. As part of this migration I also have removed several smaller disks and replaced them with larger ones (the total number of disks decreased, if that matters). I've been using an LSI SAS2308 with two breakout cables connected without issue up to this physical migration. Unraid boots back up as expected but is only "seeing" direct SATA connected drives and not anything through the LSI card (however the card itself shows up in System Devices but with nothing attached). Nothing else changed with the build and everything got connected back up just how it was before, except for the drives. 1. Could disconnecting and reconnecting the LSI card reset some sort of IT-mode settings? 2. Would the card or Unraid be unhappy that the drives all got shuffled around and some old ones aren't there anymore (keep in mind I'm not starting the array, I'm aware of mapping serial numbers before doing so)? 3. Is there a way to verify IT/passthrough mode via the Unraid terminal? Any other ideas would be appreciated!

I haven't tried this specifically with Storj, but with other duplicate dockers as long as you name them differently on install it will just create another copy - the app data doesn't know or care about other containers that are installed. The only problem would potentially be on the Storj side, with having to route two separate nodes off the same public IP. I'd also be curious if 2x 4TB nodes generates more income/usage than 1x 8TB node.

You’ll be fine. The polling for availability is over a long period. Server updates won’t be noticed.

I'm noticing that this container is keeping my disks spun up a lot. What settings are people using for their Storj share? Is there anything I should be adjusting if using Folder Caching?

Hi, trying my best to figure this out, but finally deciding to post here. Still occurring in 6.10.2. I have a pretty old Supermicro X8 with an AMD RX 560 installed. I have radeontop installed, as well as this plugin. I'm able to pass through the card to a Windows VM and use it fine. However, inside Unraid (both via this plugin's dashboard and radeontop in console) show 100% usage across the board. Nothing I'm aware of is using the card, so I'm guessing it's not accurate. Posting some screenshots here of configs. My only clue is when running radeontop -d - -l 1 I get the following output: Failed to find DRM devices: error 2 (No such file or directory) Failed to open DRM node, no VRAM support. Dumping to -, line limit 1. 1654535060.478295: bus 02, gpu 100.00%, ee 100.00%, vgt 100.00%, ta 100.00%, sx 100.00%, sh 100.00%, spi 100.00%, sc 100.00%, pa 100.00%, db 100.00%, cb 100.00% Any ideas?

I'd like to learn more about Graceful Exits. I have everything up and running, but there are still chances of gaps in availability when the container updates, I restart my server, etc. What does "leaving a Satellite" (Storj's words) mean for this docker? Should I avoid auto-updating this container and run a Graceful Exit each time it goes offline? Or is there a time limit that I need to re-appear online before I'm considered "offline"?

Thanks, this seems to be working well. I didn't realize there was another image to try!

Can you share a bit more about how you have multiple nodes from one instance? I have been following this thread and have five instances and five nodes running successfully, but I would love to only deal with one container. I have one share for Storj, and in it five main directories (Node_1, Node_2, ...). Then in docker I have my containers (Storj1, Storj2, ... ) each with successive ports (4 per container, so port 4000-4019 are all being used). Obviously each directory has its own config and node information. How are you getting around the seemingly 1:1 mapping of a Storj node and a docker container?

Can I safely run this while my array is doing other stuff (Other dockers running, such as Plex and downloaders writing to/from disk), or should I shut things down so it's the only thing manipulating data?

Thanks, I specified F@H to 0,1 and Plex to 2,3 (this is a 3.07gHz i3). Let's see how that goes. Thanks for the help!

I can definitely start messing with that option as well as the config options in F@H itself, just trying to be lazy about it. Ideally someone could write a script that toggled F@H's pause/fold state based on what is considered "idle" for unRAID, meaning no activity on disks or docker images. That would offer maximum benefit for both unRAID performance and F@H productivity.

+1 as well

https://fah-web.stanford.edu/projects/FAHClient/wiki/ClientUserGuide