robinh

Ye you have to change your postgres repository to: tensorchord/pgvecto-rs:pg14-v0.1.11 After that you won't get the 'Unable to connect to the database' error it was part of the changelog.

In that case they might have solved the issue since on sunday the following version was getting installed: mergerfs 2.33.5-22-g629806e. I will reboot my Unraid machine later this week to reinstall the 6.11.3. Did install it last weekend but was suspecting the release as root-cause for the issues with mergerfs so reverted it back to 6.11.2.

The release on the github is from april indeed but the docker is pulling the master which is recently updated ( 4 days ago).

Are you sure the mount script is not using docker to build the mergerfs? and then place the compiled version to /bin on your host? in the posted examples in this topic you see the following happening in the mount script. Maybe you were not aware it was using docker to build the mergerfs file. # Build mergerfs binary echo "$(date "+%d.%m.%Y %T") INFO: Mergerfs not installed - installing now." mkdir -p /mnt/user/appdata/other/rclone/mergerfs docker run -v /mnt/user/appdata/other/rclone/mergerfs:/build --rm trapexit/mergerfs-static-build mv /mnt/user/appdata/other/rclone/mergerfs/mergerfs /bin

Sorry I didn't document all my steps but according to my history file it should be something like this: # Pull the docker image docker pull trapexit/mergerfs-static-build:latest # list docker images - check for image id of the mergerfs-static-build sudo docker images # Go into container mergerfs-static-build sudo docker run -it <your image ID> /bin/sh # Edit the build-mergerfs file edit the build-mergerfs # Add the checkout to use a certain version github # add after cd mergerfs git checkout d1762b2bac67fbd076d4cca0ffb2b81f91933f63 # save file # exit the container by using the exit command # Saving your settings sudo docker commit [CONTAINER_ID] [new_image_name] e.g. docker commit 1015996dd4ee test-mergerfsstat # Start new container docker run -v /mnt/user/appdata/other/rclone/mergerfs:/build --rm test-mergerfsstat /tmp/build-mergerfs #moving the compiled version to /bin mv /mnt/user/appdata/other/rclone/mergerfs/mergerfs /bin Hopefully this makes it a bit clear.

Be aware that there seems to be a bug in the latest version of mergerfs on the github page. I've noticed it after I did reboot my unraid machine and afterwards the mergerfs was crashing everytime. The crashes did occur after write events. In my dmesg logging: [ 467.808897] mergerfs[7466]: segfault at 0 ip 0000000000000000 sp 0000147fb0e0e1a8 error 14 [ 467.808921] Code: Unable to access opcode bytes at RIP 0xffffffffffffffd6. The only way to get the mounts working again was using the unmount and mounting script, but as soon there was a write event the issue occured directly again (0 kb written files). I've temporary solved it by edditing the mergerfs-static-build image so it wouldn't pull the latest version of mergerfs from github. Instead I'm using now the 'd1762b2bac67fbd076d4cca0ffb2b81f91933f63' version from 7 aug. And that seems to be working again after copying the mergerfs to /bin :-) Not working mergerfs version is: mergerfs version: 2.33.5-22-g629806e Working version is: mergerfs version: 2.33.5

I also don't get it which platform I could use since im using a Intel i3-9100 (Coffee Lake). its not compatible for my CPU? Edit got it working by using the following instructions: https://xpenology.com/forum/topic/62031-how-to-install-xpenology-dva3221-on-unraid/#comment-282630

Had the same issue and the nginx reload solved it for me aswell.

As far I know docker scan is not available on Unraid.

Any news on this issue? I also always have unclean shutdowns. Found this post and figured out that this was working for me aswell umount -l /dev/loopX. But there most be a clean way to do a reboot.