3 minutes ago, spl147 said:

edit the container, click in the right corner where it says basic view, so it will say advanced, scroll down to where it says extra parameters and put --device /dev/sr0 --device /dev/sg1, then click apply

it has been talked about here https://github.com/jlesage/docker-makemkv/issues/82

 

suprosingly enabling privileged mode, makes it work. hopefully with a container update where privileged mode would be enabled. also i hope that @Djoss updates the docker with a field for adding disc drive (with --device /dev/sr0 being default value maybe) in the basic view. overall it works well.

On 9/15/2021 at 4:34 PM, simcole said:

I have had several movies recently that wouldn't rip via GUI but do via the command line above.  Specifically Princess Bride DVD and Black Hawk Down BluRay.  Any idea why that is?

What is the issue exactly with the GUI?

2 hours ago, Goldmaster said:

it has been talked about here https://github.com/jlesage/docker-makemkv/issues/82

 

suprosingly enabling privileged mode, makes it work. hopefully with a container update where privileged mode would be enabled. also i hope that @Djoss updates the docker with a field for adding disc drive (with --device /dev/sr0 being default value maybe) in the basic view. overall it works well.

 

I think the issue you referenced on GitHub is not the same thing.

 

Also, privileged mode should not be required for MakeMKV to work when correct Linux devices are exposed to the container.  It's generally a bad idea (from a security standpoint) to enable privileged mode.

 

Finally, it's difficult to provide a template with Linux devices exposed by default, since the devices to use are not the same for every setup.  We don't want to provide default settings that would make the container creation to fail...

11 hours ago, Djoss said:

 

I think the issue you referenced on GitHub is not the same thing.

 

Also, privileged mode should not be required for MakeMKV to work when correct Linux devices are exposed to the container.  It's generally a bad idea (from a security standpoint) to enable privileged mode.

 

Finally, it's difficult to provide a template with Linux devices exposed by default, since the devices to use are not the same for every setup.  We don't want to provide default settings that would make the container creation to fail...

I totally agree. Perhaps a more detailed explanation in the container for device setup.

 

something like this perhaps:

 

NOTE: For the container to have access to your optical drive(s), you need to

add them to your containers *Extra Parameters* line. An optical drive is represented by two Linux device files: /dev/srX and /dev/sgY. For optimal performance, the container needs both of them. This is done by clicking the button on the top right labeled Basic View to switch to Advanced View, this will expose the *Extra Parameters* line. To determine the right devices to use, start the container and look at its log. Then add the Devices listed to the Extra Parameters Line.

 

Example Log Output:

[cont-init.d] 95-check-optical-drive.sh: executing...

[cont-init.d] 95-check-optical-drive.sh: looking for usable optical drives...

[cont-init.d] 95-check-optical-drive.sh: found optical drive [/dev/sr0, /dev/sg1], group 19.

[cont-init.d] 95-check-optical-drive.sh: exited 0.

 

then add as follows:

--device /dev/sr0 --device /dev/sg1

Edited September 21, 20214 yr by spl147

 

Edited September 21, 20214 yr by spl147

Hey there,

 

I have a problem with the docker. Since yesterday the docker doesnt start correctly and I cant reinstall it without having the problems again. I did remove the stuff in the appdata folder, had to do a manual reset on the machine to shut the running docker down, it didnt respond to kill etc.

There is no log output, even when starting and then crashing, there is nothing. The syslog is empty too. 

Unfortunatly I dont understand the docker from binhex, I dont get it to run with a drive etc.

 

The only thing I did change was removing the drives from the machine and connect them via a sata to usb3 cable. 3 drives at all. Every drive did work nicely. 

 

Any help appreciated

3 hours ago, masterdot said:

Hey there,

 

I have a problem with the docker. Since yesterday the docker doesnt start correctly and I cant reinstall it without having the problems again. I did remove the stuff in the appdata folder, had to do a manual reset on the machine to shut the running docker down, it didnt respond to kill etc.

There is no log output, even when starting and then crashing, there is nothing. The syslog is empty too. 

Unfortunatly I dont understand the docker from binhex, I dont get it to run with a drive etc.

 

The only thing I did change was removing the drives from the machine and connect them via a sata to usb3 cable. 3 drives at all. Every drive did work nicely. 

 

Any help appreciated

When you re-installed the container, did you keep default settings ?

No, I did reconfigure it. The devices, the output, registration and storage path. The appdata was changed too. But I did try default config too. 

31 minutes ago, masterdot said:

No, I did reconfigure it. The devices, the output, registration and storage path. The appdata was changed too. But I did try default config too. 

 

Default settings should definitely work.  Can you try again to re-create the container?  Does unRAID report any failure after the creation ?  If not, can you check container's log with this command:

docker logs MakeMKV

 

Log says:

Error response from daemon: can not get logs from container which is dead or marked for removal

 

The docker is shown as running, but no gui is accessible and I cant do anything with that. I dont know why this did happen, it did run without any trouble before... 'til yesterday...

Edited October 1, 20214 yr by masterdot
added picture

Hey there @Djoss

 

Meanwhile I did create a brand new docker image (which did work better than I expected) but the makemkv container is making problems again. The default variant did start, but after adding the devices parameter it did cause problems again. 

In the logfiles there is nothing that may help to find the issue. Its really weired, the rest is running fine and not causing any problem on starting and stopping. Maybe I did run into a issue with unraids docker... 

If you need anything, just tell me what and Ill try my best to deliver it

... OMG... I did find the error... Its not your fault, the USB to SATA thing does cause such big problems. I have three of them and one does cause the problems... Onto that defective converter I did mess up with the declaration of the devices, wrong format etc. Think Im a lot more tired than I thought... Sorry for the trouble, maybe you can ease things up with declaration of the devices? Ill keep an eye on that... Never would have thought that a USB device can cause such a problem with a docker container... 

Hey @Djoss

 

I'm really sorry for causing such a trouble. I did recieve today a new sata controller and now everything is working like expected. 

All three usb sata devices made problem AND the sata controller too... 

I don't believe it, but ok. Crisis is over and everything is working again like expected. 

I did switch from usb to a adapter cable, sata ito esata and run the drives on a own power supply. 

But it's really funny, I didn't expect a "Server Error" on docker start to be caused from a faulty sata controller... Unbelievable

Is there anything specific I need to change in order to get maximum speed with it?

I'm using two Seagate IronWolf 530 NVMe SSDs (PCIe 4.0) in raid 0 and only getting around 300MB/s. I tested the same with those two SSDs in separate mode (reading from one and writing to the other one) and I was getting the exact same performance. With these drives I was expecting the speed in the GB/s territory. 

App data of  that Docker is stored on two 510 NVMe SSDs in raid 1. 

On 11/3/2021 at 2:35 PM, Sledgehamma said:

Is there anything specific I need to change in order to get maximum speed with it?

I'm using two Seagate IronWolf 530 NVMe SSDs (PCIe 4.0) in raid 0 and only getting around 300MB/s. I tested the same with those two SSDs in separate mode (reading from one and writing to the other one) and I was getting the exact same performance. With these drives I was expecting the speed in the GB/s territory. 

App data of  that Docker is stored on two 510 NVMe SSDs in raid 1. 

Is your question generic or specific to MakeMKV?  Because when using MakeMKV, I guess the bottleneck is not the SSDs, but your optical drive.

MAX read speed for a optical drive for bluray/uhd is 6x or 27MB/s

Im reading the file off an NVMe SSD and not an optical drive. 

 

I also used /mnt/cache/XX rather than /mnt/user/share

Edited November 8, 20214 yr by Sledgehamma

5 hours ago, Sledgehamma said:

Im reading the file off an NVMe SSD and not an optical drive. 

 

I also used /mnt/cache/XX rather than /mnt/user/share

i am using a Samsung 970 EVO plus reading from and writing to the same drive i hit 450MB/s easy

 

real world numbers vs advertised numbers are very very different

22 hours ago, spl147 said:

i am using a Samsung 970 EVO plus reading from and writing to the same drive i hit 450MB/s easy

 

real world numbers vs advertised numbers are very very different

 

Yes, that is true but 300 MB/s is a lot off plus you are also seeing faster speeds than I am, even with PCIe 3.0.  And the fact that the remuxing speeds with raid0 vs single are identical for me make it look like there is a bottleneck somewhere. 

On another forum somebody used W10 with two separate SSDs and is getting around 600-700MB/s. With MKTVToolnix and with tsMuxer around 1.5GBps. 

 

Edited November 9, 20214 yr by Sledgehamma

Since the last 1-2 month this container always crashes as soon an ISO requires access to /usr/bin/java. In that situation I always switched to my Windows VM that has that same MakeMKV version running. This worked for the same ISO.

 

Investigating further I found out, that my Windows VM has no Java installed. So this seems to do the trick: No Java. Now I did the following to get the MakeMKV container running again without Java support:

 

/tmp # ls -l /usr/bin/java
lrwxrwxrwx    1 root     root            35 Nov  7 15:21 /usr/bin/java -> ../lib/jvm/default-jvm/jre/bin/java

/tmp # ls -l /usr/lib/jvm/default-jvm/jre/bin/java
-rwxr-xr-x    1 root     root         14360 Feb 15  2021 /usr/lib/jvm/default-jvm/jre/bin/java

/tmp # /usr/lib/jvm/default-jvm/jre/bin/java -version
openjdk version "1.8.0_275"
OpenJDK Runtime Environment (IcedTea 3.17.1) (Alpine 8.275.01-r0)
OpenJDK 64-Bit Server VM (build 25.275-b01, mixed mode)

/usr/lib/jvm/java-1.8-openjdk/jre/bin # mv java java.sav

 

I don't know why, but for what I do with content, Java is not required. So my question is:

 

How can I make that renaming/overwriting persistent between container restarts. Would it help to add a path mapping pointing to an empty /usr/lib/jvm?

 

Thanks in advance.

 

On 11/9/2021 at 3:01 PM, Sledgehamma said:

 

Yes, that is true but 300 MB/s is a lot off plus you are also seeing faster speeds than I am, even with PCIe 3.0.  And the fact that the remuxing speeds with raid0 vs single are identical for me make it look like there is a bottleneck somewhere. 

On another forum somebody used W10 with two separate SSDs and is getting around 600-700MB/s. With MKTVToolnix and with tsMuxer around 1.5GBps. 

 

I don't expect any performance change between the container and the host (unRAID).

 

Did you perform any disk performance test on unRAID itself ?

5 hours ago, hawihoney said:

Since the last 1-2 month this container always crashes as soon an ISO requires access to /usr/bin/java. In that situation I always switched to my Windows VM that has that same MakeMKV version running. This worked for the same ISO.

 

Investigating further I found out, that my Windows VM has no Java installed. So this seems to do the trick: No Java. Now I did the following to get the MakeMKV container running again without Java support:

 

/tmp # ls -l /usr/bin/java
lrwxrwxrwx    1 root     root            35 Nov  7 15:21 /usr/bin/java -> ../lib/jvm/default-jvm/jre/bin/java

/tmp # ls -l /usr/lib/jvm/default-jvm/jre/bin/java
-rwxr-xr-x    1 root     root         14360 Feb 15  2021 /usr/lib/jvm/default-jvm/jre/bin/java

/tmp # /usr/lib/jvm/default-jvm/jre/bin/java -version
openjdk version "1.8.0_275"
OpenJDK Runtime Environment (IcedTea 3.17.1) (Alpine 8.275.01-r0)
OpenJDK 64-Bit Server VM (build 25.275-b01, mixed mode)

/usr/lib/jvm/java-1.8-openjdk/jre/bin # mv java java.sav

 

I don't know why, but for what I do with content, Java is not required. So my question is:

 

How can I make that renaming/overwriting persistent between container restarts. Would it help to add a path mapping pointing to an empty /usr/lib/jvm?

 

Thanks in advance.

 

 

Last versions of MakeMVV definitely fixed these crashes (at least for me).

Are you still getting crashes when running the latest Docker image file ?

49 minutes ago, Djoss said:

Are you still getting crashes when running the latest Docker image file ?

 

Yes, definitely. Last time was this morning.

 

8 hours ago, Djoss said:

I don't expect any performance change between the container and the host (unRAID).

 

Did you perform any disk performance test on unRAID itself ?

Not sure whats changed (maybe a reboot was necessary for the file patch change?) but now I'm getting 1Gb/s