xtrips

Good guess. I connected the cache SSD drive to the new PCI-E controller with an old SATA cable. I now used another SATA 6Gb/s cable I had around and it seems to work fine. I should have known better, being a professional in audio-video systems I know the importance of HDMI cable versions.

Ok. I will try that. Thank you both

Hello, As a first step I upgraded my cache drive from SSD to a bigger SSD. I messed around with the docker image file so I had to reinstall all my apps, but in the end it worked fine. Then I installed a PCI-E to SATA expansion card to reach higher speeds for the cache drive. At first Unraid quickly recognized the cache drive and assigned it to its role. The apps went up so I thought it was ok. But then the failures came. Unable to update docker apps and Sabnzbd unable to complete downloads. Sometimes saying it is unable to write sometimes writing some 40 Gb of files and all of a sudden all the files already written are cleaned up so the job is unable to complete. I got this message while trying to update Sabnzbd: Error: open /var/lib/docker/tmp/GetImageBlob522948715: read-only file system Any idea? novanas2-diagnostics-20190723-0843.zip

as i said i can see the missing drive with the grayed out format option, but under Parity drives it doesnt show in the list for assignment novanas2-diagnostics-20190714-1307.zip

Hello, what the hell hell happened this morning? I woke up this morning with a running Unraid but no parity drive. The interface shows up an option to format that disk but the Format button is grayed out. I attached the log of that drive. I stopped the Array for the moment.

Hello. Pressing Detect for Autofan has no effect. What are the conditions to be met for this to work?

Hello. I replaced my cache drive and added a faster SATA controller which brought it to 6Gb/s speed. I was wondering, I have a second SATA port available on the controller, will I gain anything by porting my Parity drive to that higher speed port?

Hello. The cache disk swap went well. Importing the .dockers.img file in the end raised alerts that it might be unstable and the recommendation was to build it anew, which I did. Surprisingly reinstalling my apps found the old config and everything returned back to where it was. The only hiccup was and still is the appdata folder. Remember I had to change the setting from prefer to yes, and that moved the appdata from the cache disk somewhere else (?). Now after everything is restored I get this warning: WARNINGS FOUND SUGGESTED FIX Share appdata set to cache-only, but files / folders exist on the array You should change the share's settings appropriately or use the dolphin / krusader docker applications to move the offending files accordingly. Note that there are some valid use cases for a set up like this. In particular: THIS Move it from where? I set it to prefer again but the warning stays. I don't see it in the array.

This is what the help file says: You must specify an image file for Docker. The system will automatically create this file when the Docker service is first started. It is recommended to create this image file outside the array, e.g. on the Cache pool. For best performance SSD devices are preferred. The dockers service is stopped. How does the file remains visible then?

I could simply copy the dockers.img file to media and after I make the swap put it back on cache. And only after launch back the dockers. No?

I think I set it there. I thought it would be wise, more reliable or accessible, or faster for the dockers to load from there. I guess I was wrong since it loads in memory anyway. And no, no system share.

Exactly. The Docker vDisk Location is /mnt/cache/.docker.img Should I set it at /mnt/user/ ..... ?

Just an empty backup folder and the 20 Gb .docker.img file. I could relaunch the dockers and set the image file to be on the media. Is that a good thing to do?

No, the Docker icon has disappeared. The appdata was set to cache "prefer". I now et it to yes. Is that what you meant? Update: the appdata moved, What remains is a 20 Gb .docker.img file

Did you mean this? After I stopped the Docker service, and launched the mover, there still the appdata folder and the .docker.img file on my cache drive. Did I miss something?

Ahhh ok [emoji2] Thanks Sent from my iPad using Tapatalk Pro

I didn’t get the “set all shares to cache yes”. What does that do? How does it not write to cache then? Sent from my iPad using Tapatalk Pro

I would like to replace my SSD cache drive with a bigger SSD drive. I thought about launching the mover first. Than shutdown and clone my cache drive to the new one. And finally put the new one back in place. Is that ok? Any suggestions? Sent from my iPad using Tapatalk Pro

Order placed. Thank you for your help.

I take it that you imply that in the end I will in fact get the full 6Gb speed from an SSD drive with that solution, right? Pardon me for asking but do you assert that from experience or just based on the theory? Sent from my iPad using Tapatalk Pro

Hello, With downloads becoming bigger and bigger my cache drive speed has become a bottleneck. To be more precise my motherboard's SATA port speed is limited to 3Gb/s. I am not willing to make a complete change of my motherboard at the moment, but I found many hardware solutions that could make use on my PCIe 1.0 or PCIe 2.0 available slots and convert them to a SATA 6Gb/s port. They are cheap and if they work they would be a instant gain in speed and efficiency for my cache drive duties. So the question is will Unraid recognize these extension cards to their full potential?

Thanks Any chance the developer of the docker could step in and provide a definitive answer? Sent from my iPhone using Tapatalk Pro

But 240 Gb is way over 2 or even 3 of this big downloads in sequence. Isn't there a way to force the mover right after an unpack job? or even force Sabnzbd to write the downloaded files to cache BUT unpack straight to the Raid ?

Is there any configurations available in Sabnzbd that pertains to its behavior with the Cache Disk? If so, what is the best practice? Once upon a time, when the biggest downloads were about 40 Gb every thing was smooth. Nowadays with downloads averaging 60Gb or 70 Gb or even more I always get these alerts about the cache disk being almost full, or another exemple which I think is also connected to the problem, the Direct Unpack doing its job almost until the end and then every thing unpacked is wiped out and the whole verify then unpack process restarts from the beginning. BTW my cache disk is a 240 Gb SSD. Thanks

Why is the Listening port is always Closed no matter what I try on my router? I opened for TCP on different ports. I tried as bridge then as host. Nothing helps. Other Docker apps works correctly with a forwarded port.