I have a really noob questions around Convoy, which I hope somebody might be able to shed some light on.
I have a fairly solid understanding of how Docker volumes work, and have a rudimentary understanding of the concepts behind the Union File System (UFS). I’ve taken a read through the Convoy README and understand what Convoy does — it looks fantastic BTW!
Where I’m feeling a bit out of my depth is understanding how the different backend implementations work (Device Mapper, NFS etc) in relation to Convoy, and how I should be configuring Convoy. My experience of working with volumes so far is just to let Docker manage volumes in it’s default way — I’ve never had to delve in to the underlying drivers.
Would somebody be able to shed some light on how I should be configuring the Convoy daemon with a filesystem before I try to install and test it?
Currently Convoy supports ext4 by default for Device Mapper and EBS driver, but you can reformat the volume(see DriverInfo/Device for block device when run convoy inspect of the volume) using different FS, as long as default mount command can mount them on the host(try out convoy mount after you format it, and remember to convoy umount). As you see, Device Mapper and EBS are block level drivers, rather than filesystem level. Though we haven’t tested any kind of unionfs yet, they should just works fine. So if there are issues please let us know.
I researched a bit, found out that seems overlayfs didn’t play well with NFS, so probably your best chance is to try out our Device Mapper driver with any union fs as your choice, since EBS would be more expensive to test with.
This is all very useful. I think I will try to put some time aside to get a bit more of an understanding about how the drivers work — this is all an area where m knowledge is a little thin and I want to know a bit more about what I’m doing before I start breaking things with Convoy!
