We don’t directly support Render but if a new deploy is resetting the data it sounds like the disk is not being configured correctly under /home/node/.n8n so it would be worth checking that first.
The ELIFCYCLE error could be a couple of things so having more output would be useful there but it sounds like the instance could be in a bad place.
Disk configuration
Configure specifications for your disks. If you’ve enabled scaling, configurations will apply to all disks that are created for each instance.
Mount path /opy/render/.n8n
The absolute path to mount this disk. Cannot be the root (/) directory.
Size 1GB
You can increase the size later, but you can’t decrease it. We recommend starting with the lowest value that serves your use case.
GB
That I don’t know as I don’t know the Render platform, I would say rename the path and see what happens but don’t forget that any workflows you have now will be in the database in the image so you will lose them. It might be worth exporting them first so you can add them again once you have the storage in place.