Note - The image encoder that requires server version 4.1 or higher.
When images are uploaded to the ftrack server they need to be converted and updated with additional metadata to allow viewing them in the browser. This happens automatically for cloud hosted accounts but is not built into the on-prem installation of ftrack. Instead uploaded images will show up without a thumbnail and it is not possible to view them directly in the browser.
(ftrack 4.1 and later) To have images converted automatically an additional image conversion service needs to be running on the server and the ftrack server needs to be configured to use it.
Configure the ftrack server to use the internal image conversion service by adding this to the default section in the ftrack.ini file:
ftrack.image_conversion_service = internal
Then restart the ftrack service.
Once the ftrack service is restarted, try uploading an image such as a jpeg or png to the server. It should now show up as being encoded and a job should be in progress in the job section in the top toolbar.
Docker is not fully supported on Centos 6, but can still be installed like described below. Our recommendation is however to run the image service on another server where the latest version of docker can be installed.
To install the docker service on centos 6 the following can be used:
yum install epel-release
yum install https://get.docker.com/rpm/1.7.0/centos-6/RPMS/x86_64/docker-engine-1.7.0-1.el6.x86_64.rpm
chkconfig docker on
service docker start
Then we start the image conversion service like this:
docker run --name "image-converter1" -d --net=host -e "AMQP_HOST=<amqp://guest:guest@localhost:5672/%2f>" -e "AMQP_QUEUE=ftrack-image-conversion" --restart always ftrackdocker/image-encoder:latest
Once the docker image has been pulled and is started, check that it is running like this:
If the container is running then check back in ftrack and the job should now have finished.
jpg, jpeg, png, tiff, tif, bmp, pdf, psd, gif, dpx, exr, svg
Some formats will layers and transparency may be incorrectly presented.
In case the encoding is not working, start by looking at the log for the container, by entering the following command:
This will display the id and name of the container. So now enter:
docker logs <idornameofcontainer>
Get version of docker: